Employment Div., Dept. of Human Resources of Ore. v. Smith

*874 Justice Scalia delivered the opinion of the Court. This case requires us to decide whether the Free Exercise Clause of the First Amendment permits the State of Oregon to include religiously inspired peyote use within the reach of its general criminal prohibition on use of that drug, and thus permits the State to deny unemployment benefits to persons dismissed from their jobs because of such religiously inspired use. I Oregon law prohibits the knowing or intentional possession of a “controlled substance” unless the substance has been prescribed by a medical practitioner. Ore. Rev. Stat. §475.992(4) (1987). The law defines “controlled substance” as a drug classified in Schedules I through V of the Federal Controlled Substances Act, 21 U. S. C. §§811-812, as modified by the State Board of Pharmacy. Ore. Rev. Stat. §475.005(6) (1987). Persons who violate this provision by possessing a controlled substance listed on Schedule I are “guilty of a Class B felony.” §475.992(4)(a). As compiled by the State Board of Pharmacy under its statutory authority, see §475.035, Schedule I contains the drug peyote, a hallucinogen derived from the plant Lophophora ivilliamsii Lemaire. Ore. Admin. Rule 855-80-021(3)(s) (1988). Respondents Alfred Smith and Galen Black (hereinafter respondents) were fired from their jobs with a private drug rehabilitation organization because they ingested peyote for sacramental purposes at a ceremony of the Native American Church, of which both are members. When respondents applied to petitioner Employment Division (hereinafter petitioner) for unemployment compensation, they were determined to be ineligible for benefits because they had been discharged for work-related “misconduct.” The Oregon Court of Appeals reversed that determination, holding that the denial of benefits violated respondents’ free exercise rights under the First Amendment, *875 On appeal to the Oregon Supreme Court, petitioner argued that the denial of benefits was permissible because respondents’ consumption of peyote was a crime under Oregon law. The Oregon Supreme Court reasoned, however, that the criminality of respondents’ peyote use was irrelevant to resolution of their constitutional claim — since the purpose of the “misconduct” provision under which respondents had been disqualified was not to enforce the State’s criminal laws but to preserve the financial integrity of the compensation fund, and since that purpose was inadequate to justify the burden that disqualification imposed on respondents’ religious practice. Citing our decisions in Sherbert v. Verner, 374 U. S. 398 (1963), and Thomas v. Review Bd. of Indiana Employment Security Div., 450 U. S. 707 (1981), the court concluded that respondents were entitled to payment of unemployment benefits. Smith v. Employment Div., Dept. of Human Resources, 301 Ore. 209, 217-219, 721 P. 2d 445, 449-450 (1986). We granted certiorari. 480 U. S. 916 (1987). Before this Court in 1987, petitioner continued to maintain that the illegality of respondents’ peyote consumption was relevant to their constitutional claim. We agreed, concluding that “if a State has prohibited through its criminal laws certain kinds of religiously motivated conduct without violating the First Amendment, it certainly follows that it may impose the lesser burden of denying unemployment compensation benefits to persons who engage in that conduct.” Employment Div., Dept. of Human Resources of Oregon v. Smith, 485 U. S. 660, 670 (1988) (Smith I). We noted, however, that the Oregon Supreme Court had not decided whether respondents’ sacramental use of peyote was in fact proscribed by Oregon’s controlled substance law, and that this issue was a matter of dispute between the parties. Being “uncertain about the legality of the religious use of peyote in Oregon,” we determined that it would not be “appropriate for us to decide whether the practice is protected by the Federal Constitution.” Id., at 673. Accordingly, we *876 vacated the judgment of the Oregon Supreme Court and remanded for further proceedings. Id., at 674. On remand, the Oregon Supreme Court held that respondents’ religiously inspired use of peyote fell within the prohibition of the Oregon statute, which “makes no exception for the sacramental use” of the drug. 307 Ore. 68, 72-73, 763 P. 2d 146, 148 (1988). It then considered whether that prohibition was valid under the Free Exercise Clause, and concluded that it was not. The court therefore reaffirmed its previous ruling that the State could not deny unemployment benefits to respondents for having engaged in that practice. We again granted certiorari. 489 U. S. 1077 (1989). II Respondents’ claim for relief rests on our decisions in Sherbert v. Verner, supra, Thomas v. Review Bd. of Indiana Employment Security Div., supra, and Hobbie v. Unemployment Appeals Comm’n of Florida, 480 U. S. 136 (1987), in which we held that a State could not condition the availability of unemployment insurance on an individual’s willingness to forgo conduct required by his religion. As we observed in Smith I, however, the conduct at issue in those cases was not prohibited by law. We held that distinction to be critical, for “if Oregon does prohibit the religious use of peyote, and if that prohibition is consistent with the Federal Constitution, there is no federal right to engage in that conduct in Oregon,” and “the State is free to withhold unemployment compensation from respondents for engaging in work-related misconduct, despite its religious motivation.” 485 U. S., at 672. Now that the Oregon Supreme Court has confirmed that Oregon does prohibit the religious use of peyote, we proceed to consider whether that prohibition is permissible under the Free Exercise Clause. A The Free Exercise Clause of the First Amendment, which has been made applicable to the States by incorporation into *877 the Fourteenth Amendment, see Cantwell v. Connecticut, 310 U. S. 296, 303 (1940), provides that “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof. ...” U. S. Const., Arndt. 1 (emphasis added). The free exercise of religion means, first and foremost, the right to believe and profess whatever religious doctrine one desires. Thus, the First Amendment obviously excludes all “governmental regulation of religious beliefs as such.” Sherbert v. Verner, supra, at 402. The government may not compel affirmation of religious belief, see Torcaso v. Watkins, 367 U. S. 488 (1961), punish the expression of religious doctrines it believes to be false, United States v. Ballard, 322 U. S. 78, 86-88 (1944), impose special disabilities on the basis of religious views or religious status, see McDaniel v. Paty, 435 U. S. 618 (1978); Fowler v. Rhode Island, 345 U. S. 67, 69 (1953); cf. Larson v. Valente, 456 U. S. 228, 245 (1982), or lend its power to one or the other side in controversies over religious authority or dogma, see Presbyterian Church in U. S. v. Mary Elizabeth Blue Hull Memorial Presbyterian Church, 393 U. S. 440, 445-452 (1969); Kedroff v. St. Nicholas Cathedral, 344 U. S. 94, 95-119 (1952); Serbian Eastern Orthodox Diocese v. Milivojevich, 426 U. S. 696, 708-725 (1976). But the “exercise of religion” often involves not only belief and profession but the performance of (or abstention from) physical acts: assembling with others for a worship service, participating in sacramental use of bread and wine, proselytizing, abstaining from certain foods or certain modes of transportation. It would be true, we think (though no case of ours has involved the point), that a State would be “prohibiting the free exercise [of religion]” if it sought to ban such acts or abstentions only when they are engaged in for religious reasons, or only because of the religious belief that they display. It would doubtless be unconstitutional, for example, to ban the casting of “statues that are to be used *878 for worship purposes,” or to prohibit bowing down before a golden calf. Respondents in the present case, however, seek to carry the meaning of “prohibiting the free exercise [of religion]” one large step further. They contend that their religious motivation for using peyote places them beyond the reach of a criminal law that is not specifically directed at their religious practice, and that is concededly constitutional as applied to those who use the drug for other reasons. They assert, in other words, that “prohibiting the free exercise [of religion]” includes requiring any individual to observe a generally applicable law that requires (or forbids) the performance of an act that his religious belief forbids (or requires). As a textual matter, we do not think the words must be given that meaning. It is no more necessary to regard the collection of a general tax, for example, as “prohibiting the free exercise [of religion]” by those citizens who believe support of organized government to be sinful, than it is to regard the same tax as “abridging the freedom ... of the press” of those publishing companies that must pay the tax as a condition of staying in business. It is a permissible reading of the text, in the one case as in the other, to say that if prohibiting the exercise of religion (or burdening the activity of printing) is not the object of the tax but merely the incidental effect of a generally applicable and otherwise valid provision, the First Amendment has not been offended. Compare Citizen Publishing Co. v. United States, 394 U. S. 131, 139 (1969) (upholding application of antitrust laws to press), with Grosjean v. American Press Co., 297 U. S. 233, 250-251 (1936) (striking down license tax applied only to newspapers with weekly circulation above a specified level); see generally Minneapolis Star & Tribune Co. v. Minnesota Comm’r of Revenue, 460 U. S. 575, 581 (1983). Our decisions reveal that the latter reading is the correct one. We have never held that an individual’s religious be *879 liefs excuse him from compliance with an otherwise valid law prohibiting conduct that the State is free to regulate. On the contrary, the record of more than a century of our free exercise jurisprudence contradicts that proposition. As described succinctly by Justice Frankfurter in Minersville School Dist. Bd. of Ed. v. Gobitis, 310 U. S. 586, 594-595 (1940): “Conscientious scruples have not, in the course of the long struggle for religious toleration, relieved the individual from obedience to a general law not aimed at the promotion or restriction of religious beliefs. The mere possession of religious convictions which contradict the relevant concerns of a political society does not relieve the citizen from the discharge of political responsibilities (footnote omitted).” We first had occasion to assert that principle in Reynolds v. United States, 98 U. S. 145 (1879), where we rejected the claim that criminal laws against polygamy could not be constitutionally applied to those whose religion commanded the practice. “Laws,” we said, “are made for the government of actions, and while they cannot interfere with mere religious belief and opinions, they may with practices. . . . Can a man excuse his practices to the contrary because of his religious belief? To permit this would be to make the professed doctrines of religious belief superior to the law of the land, and in effect to permit every citizen to become a law unto himself.” Id., at 166-167. Subsequent decisions have consistently held that the right of free exercise does not relieve an individual of the obligation to comply with a “valid and neutral law of general applicability on the ground that the law proscribes (or prescribes) conduct that his religion prescribes (or proscribes).” United States v. Lee, 455 U. S. 252, 263, n. 3 (1982) (Stevens, J., concurring in judgment); see Minersville School Dist. Bd. of Ed. v. Gobitis, supra, at 595 (collecting cases). In Prince v. Massachusetts, 321 U. S. 158 (1944), we held that a mother could be prosecuted under the child labor laws *880 for using her children to dispense literature in the streets, her religious motivation notwithstanding. We found no constitutional infirmity in “excluding [these children] from doing there what no other children may do.” Id., at 171. In Braunfeld v. Brown, 366 U. S. 599 (1961) (plurality opinion), we upheld Sunday-closing laws against the claim that they burdened the religious practices of persons whose religions compelled them to refrain from work on other days. In Gillette v. United States, 401 U. S. 437, 461 (1971), we sustained the military Selective Service System against the claim that it violated free exercise by conscripting persons who opposed a particular war on religious grounds. Our most recent decision involving a neutral, generally applicable regulatory law that compelled activity forbidden by an individual’s religion was United States v. Lee, 455 U. S., at 258-261. There, an Amish employer, on behalf of himself and his employees, sought exemption from collection and payment of Social Security taxes on the ground that the Amish faith prohibited participation in governmental support programs. We rejected the claim that an exemption was constitutionally required. There would be no way, we observed, to distinguish the Amish believer’s objection to Social Security taxes from the religious objections that others might have to the collection or use of other taxes. - “If, for example, a religious adherent believes war is a sin, and if a certain percentage of the federal budget can be identified as devoted to war-related activities, such individuals would have a similarly valid claim to be exempt from paying that percentage of the income tax. The tax system could not function if denominations were allowed to challenge the tax system because tax payments were spent in a manner that violates their religious belief.” Id., at 260. Cf. Hernandez v. Commissioner, 490 U. S. 680 (1989) (rejecting free exercise challenge to payment of income taxes alleged to make religious activities more difficult). *881 The only decisions in which we have held that the First Amendment bars application of a neutral, generally applicable law to religiously motivated action have involved not the Free Exercise Clause alone, but the Free Exercise Clause in conjunction with other constitutional protections, such as freedom of speech and of the press, see Cantwell v. Connecticut, 310 U. S., at 304-307 (invalidating a licensing system for religious and charitable solicitations under which the administrator had discretion to deny a license to any cause he deemed nonreligious); Murdock v. Pennsylvania, 319 U. S. 105 (1943) (invalidating a flat tax on solicitation as applied to the dissemination of religious ideas); Follett v. McCormick, 321 U. S. 573 (1944) (same), or the right of parents, acknowledged in Pierce v. Society of Sisters, 268 U. S. 510 (1925), to direct the education of their children, see Wisconsin v. Yoder, 406 U. S. 205 (1972) (invalidating compulsory school-attendance laws as applied to Amish parents who refused on religious grounds to send their children to school). 1 *882 Some of our cases prohibiting compelled expression, decided exclusively upon free speech grounds, have also involved freedom of religion, cf. Wooley v. Maynard, 430 U. S. 705 (1977) (invalidating compelled display of a license plate slogan that offended individual religious beliefs); West Virginia Bd. of Education v. Barnette, 319 U. S. 624 (1943) (invalidating compulsory flag salute statute challenged by religious objectors). And it is easy to envision a case in which a challenge on freedom of association grounds would likewise be reinforced by Free Exercise Clause concerns. Cf. Roberts v. United States Jaycees, 468 U. S. 609, 622 (1984) (“An individual’s freedom to speak, to worship, and to petition the government for the redress of grievances could not be vigorously protected from interference by the State [if] a correlative freedom to engage in group effort toward those ends were not also guaranteed”). The present case does not present such a hybrid situation, but a free exercise claim unconnected with any communicative activity or parental right. Respondents urge us to hold, quite simply, that when otherwise prohibitable conduct is accompanied by religious convictions, not only the convictions but the conduct itself must be free from governmental regulation. We have never held that, and decline to do so now. There being no contention that Oregon’s drug law represents an attempt to regulate religious beliefs, the communication of religious beliefs, or the raising of one’s children in those beliefs, the rule to which we have adhered ever since Reynolds plainly controls. “Our cases do not at their farthest reach support the proposition that a stance of conscientious opposition relieves an objector from any colliding duty fixed by a democratic government.” Gillette v. United States, supra, at 461. B Respondents argue that even though exemption from generally applicable criminal laws need not automatically be extended to religiously motivated actors, at least the claim for a *883 religious exemption must be evaluated under the balancing test set forth in Sherbert v. Verner, 374 U. S. 398 (1963). Under the Sherbert test, governmental actions that substantially burden a religious practice must be justified by a compelling governmental interest. See id., at 402-403; see also Hernandez v. Commissioner, 490 U. S., at 699. Applying that test we have, on three occasions, invalidated state unemployment compensation rules that conditioned the availability of benefits upon an applicant’s willingness to work under conditions forbidden by his religion. See Sherbert v. Verner, supra; Thomas v. Review Bd. of Indiana Employment Security Div., 450 U. S. 707 (1981); Hobbie v. Unemployment Appeals Comm’n of Florida, 480 U. S. 136 (1987). We have never invalidated any governmental action on the basis of the Sherbert test except the denial of unemployment compensation. Although we have sometimes purported to apply the Sherbert test in contexts other than that, we have always found the test satisfied, see United States v. Lee, 455 U. S. 252 (1982); Gillette v. United States, 401 U. S. 437 (1971). In recent years we have abstained from applying the Sherbert test (outside the unemployment compensation field) at all. In Bowen v. Roy, 476 U. S. 693 (1986), we declined to apply Sherbert analysis to a federal statutory scheme that required benefit applicants and recipients to provide them Social Security numbers. The plaintiffs in that case asserted that it would violate their religious beliefs to obtain and provide a Social Security number for their daughter. We held the statute’s application to the plaintiffs valid regardless of whether it was necessary to effectuate a compelling interest. See 476 U. S., at 699-701. In Lyng v. Northwest Indian Cemetery Protective Assn., 485 U. S. 439 (1988), we declined to apply Sherbert analysis to the Government’s logging and road construction activities on lands used for religious purposes by several Native American Tribes, even though it was undisputed that the activities “could have devastating effects on traditional Indian religious practices,” 485 U. S., at 451. *884 In Goldman v. Weinberger, 475 U. S. 503 (1986), we rejected application of the Sherbert test to military dress regulations that forbade the wearing of yarmulkes. In O'Lone v. Estate of Shabazz, 482 U. S. 342 (1987), we sustained, without mentioning the Sherbert test, a, prison’s refusal to excuse inmates from work requirements to attend worship services. Even if we were inclined to breathe into Sherbert some life beyond the unemployment compensation field, we would not apply it to require exemptions from a generally applicable criminal law. The Sherbert test, it must be recalled, was developed in a context that lent itself to individualized governmental assessment of the reasons for the relevant conduct. As a plurality of the Court noted in Roy, a distinctive feature of unemployment compensation programs is that their eligibility criteria invite consideration of the particular circumstances behind an applicant’s unemployment: “The statutory conditions [in Sherbert and Thomas] provided that a person was not eligible for unemployment compensation benefits if, ‘without good cause,’ he had quit work’ or refused available work. The ‘good cause’ standard created a mechanism for individualized exemptions.” Bowen v. Roy, supra, at 708 (opinion of Burger, C. J., joined by Powell and Rehnquist, JJ.). See also Sherbert, supra, at 401, n. 4 (reading state unemployment compensation law as allowing benefits for unemployment caused by at least some “personal reasons”). As the plurality pointed out in Roy, our decisions in the unemployment cases stand for the proposition that where the State has in place a system of individual exemptions, it may not refuse to extend that system to cases of “religious hardship” without compelling reason. Bowen v. Roy, supra, at 708. Whether or not the decisions are that limited, they at least have nothing to do with an across-the-board criminal prohibition on a particular form of conduct. Although, as noted earlier, we have sometimes used the Sherbert test to analyze free exercise challenges to such laws, see United States v. *885 Lee, supra, at 257-260; Gillette v. United States, supra, at 462, we have never applied the test to invalidate one. We conclude today that the sounder approach, and the approach in accord with the vast majority of our precedents, is to hold the test inapplicable to such challenges. The government’s ability to enforce generally applicable prohibitions of socially harmful conduct, like its ability to carry out other aspects of public policy, “cannot depend on measuring the effects of a governmental action on a religious objector’s spiritual development.” Lyng, supra, at 451. To make an individual’s obligation to obey such a law contingent upon the law’s coincidence with his religious beliefs, except where the State’s interest is “compelling” — permitting him, by virtue of his beliefs, “to become a law unto himself,” Reynolds v. United States, 98 U. S., at 167 — contradicts both constitutional tradition and common sense. 2 The “compelling government interest” requirement seems benign, because it is familiar from other fields. But using it as the standard that must be met before the government may accord different treatment on the basis of race, see, e. g., *886 Palmore v. Sidoti, 466 U. S. 429, 432 (1984), or before the government may regulate the content of speech, see, e. g., Sable Communications of California v. FCC, 492 U. S. 115, 126 (1989), is not remotely comparable to using it for the purpose asserted here. What it produces in those other fields — equality of treatment and an unrestricted flow of contending speech — are constitutional norms; what it would produce here — a private right to ignore generally applicable laws — is a constitutional anomaly. 3 Nor is it possible to limit the impact of respondents’ proposal by requiring a “compelling state interest” only when the conduct prohibited is “central” to the individual’s religion. Cf. Lyng v. Northwest Indian Cemetery Protective Assn., 485 U. S., at 474-476 (BRENNAN, J., dissenting). It is no *887 more appropriate for judges to determine the “centrality” of religious beliefs before applying a “compelling interest” test in the free exercise field, than it would be for them to determine the “importance” of ideas before applying the “compelling interest” test in the free speech field. What principle of law or logic can be brought to bear to contradict a believer’s assertion that a particular act is “central” to his personal faith? Judging the centrality of different religious practices is akin to the unacceptable “business of evaluating the relative merits of differing religious claims.” United States v. Lee, 455 U. S., at 263 n. 2 (Stevens, J., concurring). As we reaffirmed only last Term, “[i]t is not within the judicial ken to question the centrality of particular beliefs or practices to a faith, or the validity of particular litigants’ interpretations of those creeds.” Hernandez v. Commissioner, 490 U. S., at 699. Repeatedly and in many different contexts, we have warned that courts must not presume to determine the place of a particular belief in a religion or the plausibility of a religious claim. See, e. g., Thomas v. Review Bd. of Indiana Employment Security Div., 450 U. S., at 716; Presbyterian Church in U. S. v. Mary Elizabeth Blue Hull Memorial Presbyterian Church, 393 U. S., at 450; Jones v. Wolf, 443 U. S. 595, 602-606 (1979); United States v. Ballard, 322 U. S. 78, 85-87 (1944). 4 *888 If the “compelling interest” test is to be applied at all, then, it must be applied across the board, to all actions thought to be religiously commanded. Moreover, if “compelling interest” really means what it says (and watering it down here would subvert its rigor in the other fields where it is applied), many laws will not meet the test. Any society adopting such a system would be courting anarchy, but that danger increases in direct proportion to the society’s diversity of religious beliefs, and its determination to coerce or suppress none of them. Precisely because “we are a cosmopolitan nation made up of people of almost every conceivable religious preference,” Braunfeld v. Brown, 366 U. S., at 606, and precisely because we value and protect that religious divergence, we cannot afford the luxury of deeming presumptively invalid, as applied to the religious objector, every regulation of conduct that does not protect an interest of the highest order. The rule respondents favor would open the prospect of constitutionally required religious exemptions from civic obligations of almost every conceivable kind — ranging from *889 compulsory military service, see, e. g., Gillette v. United States, 401 U. S. 437 (1971), to the payment of taxes, see, e. g., United States v. Lee, supra; to health and safety regulation such as manslaughter and child neglect laws, see, e. g., Funkhouser v. State, 763 P. 2d 695 (Okla. Crim. App. 1988), compulsory vaccination laws, see, e. g., Cude v. State, 237 Ark. 927, 377 S. W. 2d 816 (1964), drug laws, see, e. g., Olsen v. Drug Enforcement Administration, 279 U. S. App. D. C. 1, 878 F. 2d 1458 (1989), and traffic laws, see Cox v. New Hampshire, 312 U. S. 569 (1941); to social welfare legislation such as minimum wage laws, see Tony and Susan Alamo Foundation v. Secretary of Labor, 471 U. S. 290 (1985), child labor laws, see Prince v. Massachusetts, 321 U. S. 158 (1944), animal cruelty laws, see, e. g., Church of the Lukumi Babalu Aye Inc. v. City of Hialeah, 723 F. Supp. 1467 (SD Fla. 1989), cf. State v. Massey, 229 N. C. 734, 51 S. E. 2d 179, appeal dism’d, 336 U. S. 942 (1949), environmental protection laws, see United States v. Little, 638 F. Supp. 337 (Mont. 1986), and laws providing for equality of opportunity for the races, see, e. g., Bob Jones University v. United States, 461 U. S. 574, 603-604 (1983). The First Amendment’s protection of religious liberty does not require this. 5 *890 Values that are protected against government interference through enshrinement in the Bill of Rights are not thereby banished from the political process. Just as a society that believes in the negative protection accorded to the press by the First Amendment is likely to enact laws that affirmatively foster the dissemination of the printed word, so also a society that believes in the negative protection accorded to religious belief can be expected to be solicitous of that value in its legislation as well. It is therefore not surprising that a number of States have made an exception to their drug laws for sacramental peyote use. See, e. g., Ariz. Rev. Stat. Ann. §§ 13 — 3402(B)(1)—(3) (1989); Colo. Rev. Stat. § 12-22-317(3) (1985); N. M. Stat. Ann. §30-31-6(D) (Supp. 1989). But to say that a nondiscriminatory religious-practice exemption is permitted, or even that it is desirable, is not to say that it is constitutionally required, and that the appropriate occasions for its creation can be discerned by the courts. It may fairly be said that leaving accommodation to the political process will place at a relative disadvantage those religious practices that are not widely engaged in; but that unavoidable consequence of democratic government must be preferred to a system in which each conscience is a law unto itself or in which judges weigh the social importance of all laws against the centrality of all religious beliefs. * * * Because respondents’ ingestion of peyote was prohibited under Oregon law, and because that prohibition is constitutional, Oregon may, consistent with the Free Exercise Clause, deny respondents unemployment compensation when their dismissal results from use of the drug. The decision of the Oregon Supreme Court is accordingly reversed. It is so ordered. *891 Justice O’Connor, with whom Justice Brennan, Justice Marshall, and Justice Blackmun join as to Parts I and II, concurring in the judgment. * Although I agree with the result the Court reaches in this case, I cannot join its opinion. In my view, today’s holding dramatically departs from well-settled First Amendment jurisprudence, appears unnecessary to resolve the question presented, and is incompatible with our Nation’s fundamental commitment to individual religious liberty. I At the outset, I note that I agree with the Court’s implicit determination that the constitutional question upon which we granted review — whether the Free Exercise Clause protects a person’s religiously motivated use of peyote from the reach of a State’s general criminal law prohibition — is properly presented in this case. As the Court recounts, respondents Alfred Smith and Galen Black (hereinafter respondents) were denied unemployment compensation benefits because their sacramental use of peyote constituted work-related “misconduct,” not because they violated Oregon’s general criminal prohibition against possession of peyote. We held, however, in Employment Div., Dept. of Human Resources of Oregon v. Smith, 485 U. S. 660 (1988) (Smith I), that whether a State may, consistent with federal law, deny unemployment compensation benefits to persons for their religious use of peyote depends on whether the State, as a matter of state law, has criminalized the underlying conduct. See id., at 670-672. The Oregon Supreme Court, on remand from this Court, concluded that “the Oregon statute against possession of controlled substances, which include peyote, makes no exception for the sacramental use of peyote.” 307 Ore. 68, 72-73, 763 P. 2d 146, 148 (1988) (footnote omitted). *892 Respondents contend that, because the Oregon Supreme Court declined to decide whether the Oregon Constitution prohibits criminal prosecution for the religious use of peyote, see id., at 73, n. 3, 763 P. 2d, at 148, n. 3, any ruling on the federal constitutional question would be premature. Respondents are of course correct that the Oregon Supreme Court may eventually decide that the Oregon Constitution requires the State to provide an exemption from its general criminal prohibition for the religious use of peyote. Such a decision would then reopen the question whether a State may nevertheless deny unemployment compensation benefits to claimants who are discharged for engaging in such conduct. As the case comes to us today, however, the Oregon Supreme Court has plainly ruled that Oregon’s prohibition against possession of controlled substances does not contain an exemption for the religious use of peyote. In light of our decision in Smith I, which makes this finding a “necessary predicate to a correct evaluation of respondents’ federal claim,” 485 U. S., at 672, the question presented and addressed is properly before the Court. II The Court today extracts from our long history of free exercise precedents the single categorical rule that “if prohibiting the exercise of religion . . . is . . . merely the incidental effect of a generally applicable and otherwise valid provision, the First Amendment has not been offended.” Ante, at 878 (citations omitted). Indeed, the Court holds that where the law is a generally applicable criminal prohibition, our usual free exercise jurisprudence does not even apply. Ante, at 884. To reach this sweeping result, however, the Court must not only give a strained reading of the First Amendment but must also disregard our consistent application of free exercise doctrine to cases involving generally applicable regulations that burden religious conduct. *893 A The Free Exercise Clause of the First Amendment commands that “Congress shall make no law . . . prohibiting the free exercise [of religion].” In Cantwell v. Connecticut, 310 U. S. 296 (1940), we held that this prohibition applies to the States by incorporation into the Fourteenth Amendment and that it categorically forbids government regulation of religious beliefs. Id., at 303. As the Court recognizes, however, the “free exercise” of religion often, if not invariably, requires the performance of (or abstention from) certain acts. Ante, at 877; cf. 3 A New English Dictionary on Historical Principles 401-402 (J. Murray ed. 1897) (defining “exercise” to include “[t]he practice and performance of rites and ceremonies, worship, etc.; the right or permission to celebrate the observances (of a religion)” and religious observances such as acts of public and private worship, preaching, and prophesying). “[B]elief and action cannot be neatly confined in logic-tight compartments.” Wisconsin v. Yoder, 406 U. S. 205, 220 (1972). Because the First Amendment does not distinguish between religious belief and religious conduct, conduct motivated by sincere religious belief, like the belief itself, must be at least presumptively protected by the Free Exercise Clause. The Court today, however, interprets the Clause to permit the government to prohibit, without justification, conduct mandated by an individual’s religious beliefs, so long as that prohibition is generally applicable. Ante, at 878. But a law that prohibits certain conduct — conduct that happens to be an act of worship for someone — manifestly does prohibit that person’s free exercise of his religion. A person who is barred from engaging in religiously motivated conduct is barred from freely exercising his religion. Moreover, that person is barred from freely exercising his religion regardless of whether the law prohibits the conduct only when engaged in for religious reasons, only by members of that religion, or by all persons. It is difficult to deny that a law that prohib *894 its religiously motivated conduct, even if the law is generally applicable, does not at least implicate First Amendment concerns. The Court responds that generally applicable laws are “one large step” removed from laws aimed at specific religious practices. Ibid. The First Amendment, however, does not distinguish between laws that are generally applicable and laws that target particular religious practices. Indeed, few States would be so naive as to enact a law directly prohibiting or burdening a religious practice as such. Our free exercise cases have all concerned generally applicable laws that had the effect of significantly burdening a religious practice. If the First Amendment is to have any vitality, it ought not be construed to cover only the extreme and hypothetical situation in which a State directly targets a religious practice. As we have noted in a slightly different context, “‘[s]uch a test has no basis in precedent and relegates a serious First Amendment value to the barest level of minimum scrutiny that the Equal Protection Clause already provides. ’ ” Hobbie v. Unemployment Appeals Comm’n of Florida, 480 U. S. 136, 141-142 (1987) (quoting Bowen v. Roy, 476 U. S. 693, 727 (1986) (O’Connor, J., concurring in part and dissenting in part)). To say that a person’s right to free exercise has been burdened, of course, does not mean that he has an absolute right to engage in the conduct. Under our established First Amendment jurisprudence, we have recognized that the freedom to act, unlike the freedom to believe, cannot be absolute. See, e. g., Cantwell, supra, at 304; Reynolds v. United States, 98 U. S. 145, 161-167 (1879). Instead, we have respected both the First Amendment’s express textual mandate and the governmental interest in regulation of conduct by requiring the government to justify any substantial burden on religiously motivated conduct by a compelling state interest and by means narrowly tailored to achieve that interest. See Hernandez v. Commissioner, 490 U. S. 680, 699 *895 (1989); Hobbie, supra, at 141; United States v. Lee, 455 U. S. 252, 257-258 (1982); Thomas v. Review Bd. of Indiana Employment Security Div., 450 U. S. 707, 718 (1981); McDaniel v. Paty, 435 U. S. 618, 626-629 (1978) (plurality opinion); Yoder, supra, at 215; Gillette v. United States, 401 U. S. 437, 462 (1971); Sherbert v. Verner, 374 U. S. 398, 403 (1963); see also Bowen v. Roy, supra, at 732 (opinion concurring in part and dissenting in part); West Virginia State Bd. of Ed. v. Barnette, 319 U. S. 624, 639 (1943). The compelling interest test effectuates the First Amendment’s command that religious liberty is an independent liberty, that it occupies a preferred position, and that the Court will not permit encroachments upon this liberty, whether direct or indirect, unless required by clear and compelling governmental interests “of the highest order,” Yoder, supra, at 215. “Only an especially important governmental interest pursued by narrowly tailored means can justify exacting a sacrifice of First Amendment freedoms as the price for an equal share of the rights, benefits, and privileges enjoyed by other citizens.” Roy, supra, at 728 (opinion concurring in part and dissenting in part). The Court attempts to support its narrow reading of the Clause by claiming that “[w]e have never held that an individual’s religious beliefs excuse him from compliance with an otherwise valid law prohibiting conduct that the State is free to regulate.” Ante, at 878-879. But as the Court later notes, as it must, in cases such as Cantwell and Yoder we have in fact interpreted the Free Exercise Clause to forbid application of a generally applicable prohibition to religiously motivated conduct. See Cantwell, supra, at 304-307; Yoder, 406 U. S., at 214-234. Indeed, in Yoder we expressely rejected the interpretation the Court now adopts: “[0]ur decisions have rejected the idea that religiously grounded conduct is always outside the protection of the Free Exercise Clause. It is true that activities of individuals, even when religiously based, are often subject *896 to regulation by the States in the exercise of their undoubted power to promote the health, safety, and general welfare, or the Federal Government in the exercise of its delegated powers. But to agree that religiously grounded conduct must often be subject to the broad police power of the State is not to deny that there are areas of conduct protected by the Free Exercise Clause of the First Amendment and thus beyond the power of the State to control, even under regulations of general applicability. . . . "... A regulation neutral on its face may, in its application, nonetheless offend the constitutional requirement for government neutrality if it unduly burdens the free exercise of religion.” Id., at 219-220 (emphasis added; citations omitted). The Court endeavors to escape from our decisions in Cant-well and Yoder by labeling them “hybrid” decisions, ante, at 892, but there is no denying that both cases expressly relied on the Free Exercise Clause, see Cantwell, 310 U. S., at 303-307; Yoder, supra, at 219-229, and that we have consistently regarded those cases as part of the mainstream of our free exercise jurisprudence. Moreover, in each of the other cases cited by the Court to support its categorical rule, ante, at 879-880, we rejected the particular constitutional claims before us only after carefully weighing the competing interests. See Prince v. Massachusetts, 321 U. S. 158, 168-170 (1944) (state interest in regulating children’s activities justifies denial of religious exemption from child labor laws); Braunfeld v. Brown, 366 U. S. 599, 608-609 (1961) (plurality opinion) (state interest in uniform day of rest justifies denial of religious exemption from Sunday closing law); Gillette, supra, at 462 (state interest in military affairs justifies denial of religious exemption from conscription laws); Lee, supra, at 258-259 (state interest in comprehensive Social Security system justifies denial of religious exemption from mandatory participation requirement). That we rejected the free exer *897 cise claims in those cases hardly calls into question the applicability of First Amendment doctrine in the first place. Indeed, it is surely unusual to judge the vitality of a constitutional doctrine by looking to the win-loss record of the plaintiffs who happen to come before us. B Respondents, of course, do not contend that their conduct is automatically immune from all governmental regulation simply because it is motivated by their sincere religious beliefs. The Court’s rejection of that argument, ante, at 882, might therefore be regarded as merely harmless dictum. Rather, respondents invoke our traditional compelling interest test to argue that the Free Exercise Clause requires the State to grant them a limited exemption from its general criminal prohibition against the possession of peyote. The Court today, however, denies them even the opportunity to make that argument, concluding that “the sounder approach.,, and the approach in accord with the vast majority of our precedents, is to hold the [compelling interest] test inapplicable to” challenges to general criminal prohibitions'. Ante, at 885. In my view, however, the essence of a. free' exercise claim, is relief from a burden imposed by government on religious’ practices or beliefs, whether the burden is imposed directly through laws that prohibit or compel specific religious practices, or indirectly through laws that, in effect, make abandonment of one’s own religion or conformity to the religious beliefs of others the price of an equal place in the civil community. As we explained in Thomas: “Where the state conditions receipt of an important benefit upon conduct proscribed by a religious faith, or where it denies such a benefit because of conduct mandated by religious belief, thereby putting substantial pressure on an adherent to modify his behavior and to violate his beliefs, a burden upon religion exists.” 450 U. S., at 717-718. *898 See also Frazee v. Illinois Dept. of Employment Security, 489 U. S. 829, 832 (1989); Hobbie, 480 U. S., at 141. A State that makes criminal an individual’s religiously motivated conduct burdens that individual’s free exercise of religion in the severest manner possible, for it “results in the choice to the individual of either abandoning his religious principle or facing criminal prosecution.” Braunfeld, supra, at 605. I would have thought it beyond argument that such laws implicate free exercise concerns. Indeed, we have never distinguished between cases in which a State conditions receipt of a benefit on conduct prohibited by religious beliefs and cases in which a State affirmatively prohibits such conduct. The Sherbert compelling interest test applies in both kinds of cases. See, e. g., Lee, 455 U. S., at 257-260 (applying Sherbert to uphold Social Security tax liability); Gillette, 401 U. S., at 462 (applying Sherbert to uphold military conscription requirement); Yoder, 406 U. S., at 215-234 (applying Sherbert to strike down criminal convictions for violation of compulsory school attendance law). As I noted in Bowen v. Roy: “The fact that the underlying dispute involves an award of benefits rather than an exaction of penalties does not grant the Government license to apply a different version of the Constitution. ... “. . . The fact that appellees seek exemption from a precondition that the Government attaches to an award of benefits does not, therefore, generate a meaningful distinction between this case and one where appellees seek an exemption from the Government’s imposition of penalties upon them.” 476 U. S., at 731-732 (opinion concurring in part and dissenting in part). See also Hobbie, supra, at 141-142; Sherbert, 374 U. S., at 404. I would reaffirm that principle today: A neutral criminal law prohibiting conduct that a State may legitimately regulate is, if anything, more burdensome than a neutral civil *899 statute placing legitimate conditions on the award of a state benefit. Legislatures, of course, have always been “left free to reach actions which were in violation of social duties or subversive of good order.” Reynolds, 98 U. S., at 164; see also Yoder, supra, at 219-220; Braunfeld, 366 U. S., at 603-604. Yet because of the close relationship between conduct and religious belief, “[i]n every case the power to regulate must be so exercised as not, in attaining a permissible end, unduly to infringe the protected freedom.” Cantwell, 310 U. S., at 304. Once it has been shown that a government regulation or criminal prohibition burdens the free exercise of religion, we have consistently asked the government to demonstrate that unbending application of its regulation to the religious objector “is essential to accomplish an overriding governmental interest,” Lee, supra, at 257-258, or represents “the least restrictive means of achieving some compelling state interest,” Thomas, supra, at 718. See, e. g., Braunfeld, supra, at 607; Sherbert, supra, at 406; Yoder, supra, at 214-215; Roy, 476 U. S., at 728-732 (opinion concurring in part and dissenting in part). To me, the sounder approach — the approach more consistent with our role as judges to decide each case on its individual merits — is to apply this test in each case to determine whether the burden on the specific plaintiffs before us is constitutionally significant and whether the particular criminal interest asserted by the State before us is compelling. Even if, as an empirical matter, a government’s criminal laws might usually serve a compelling interest in health, safety, or public order, the First Amendment at least requires a case-by-case determination of the question, sensitive to the facts of each particular claim. Cf. McDaniel, 435 U. S., at 628, n. 8 (plurality opinion) (noting application of Sherbert to general criminal prohibitions and the “delicate balancing required by our decisions in” Sherbert and Yoder). Given the range of conduct that a State might legitimately make *900 criminal, we cannot assume, merely because a law carries criminal sanctions and is generally applicable, that the First' Amendment never requires the State to grant a limited exemption for religiously motivated conduct. Moreover, we have not “rejected” or “declined to apply” the compelling interest test in our recent cases. Ante, at 883-884. Recent cases have instead affirmed that test as a fundamental part of our First Amendment doctrine. See, e. g., Hernandez, 490 U. S., at 699; Hobbie, supra, at 141— 142 (rejecting Chief Justice Burger’s suggestion in Roy, supra, at 707-708, that free exercise claims be assessed under a less rigorous “reasonable means” standard). The cases cited by the Court signal no retreat from our consistent adherence to the compelling interest test. In both Bowen v. Roy, supra, and Lyng v. Northwest Indian Cemetery Protective Assn., 485 U. S. 439 (1988), for example, we expressly distinguished Sherbert on the ground that the First Amendment does not “require the Government itself to behave in ways that the individual believes will further his or her spiritual development .... The Free Exercise Clause simply cannot be understood to require the Government to conduct its own internal affairs in ways that comport with the religious beliefs of particular citizens.” Roy, supra, at 699; see Lyng, supra, at 449. This distinction makes sense because “the Free Exercise Clause is written in terms of what the government cannot do to the individual, not in terms of what the individual can exact from the government.” Sherbert, supra, at 412 (Douglas, J., concurring). Because the case sub judice, like the other cases in which we have applied Sherbert, plainly falls into the former category, I would apply those established precedents to the facts of this case. Similarly, the other cases cited by the Court for the proposition that we have rejected application of the Sherbert test outside the unemployment compensation field, ante, at 884, are distinguishable because they arose in the narrow, specialized contexts in which we have not traditionally re *901 quired the government to justify a burden on religious conduct by articulating a compelling interest. See Goldman v. Weinberger, 475 U. S. 503, 507 (1986) (“Our review of military regulations challenged on First Amendment grounds is far more deferential than constitutional review of similar laws or regulations designed for civilian society”); O’Lone v. Estate of Shabazz, 482 U. S. 342, 349 (1987) (“[Pjrison regulations alleged to infringe constitutional rights are judged under a ‘reasonableness’ test less restrictive than that ordinarily applied to alleged infringements of fundamental constitutional rights”) (citation omitted). That we did not apply the compelling interest test in these cases says nothing about whether the test should continue to apply in paradigm free exercise cases such as the one presented here. The Court today gives no convincing reason to depart from settled First Amendment jurisprudence. There is nothing talismanic about neutral laws of general applicability or general criminal prohibitions, for laws neutral toward religion can coerce a person to violate his religious conscience or intrude upon his religious duties just as effectively as laws aimed at religion. Although the Court suggests that the compelling interest test, as applied to generally applicable laws, would result in a “constitutional anomaly,” ante, at 886, the First Amendment unequivocally makes freedom of religion, like freedom from race discrimination and freedom of speech, a “constitutional nor[m],” not an “anomaly.” Ibid. Nor would application of our established free exercise doctrine to this case necessarily be incompatible with our equal protection cases. Cf. Rogers v. Lodge, 458 U. S. 613, 618 (1982) (race-neutral law that “ ‘bears more heavily on one race than another’” may violate equal protection) (citation omitted); Castaneda v. Partida, 430 U. S. 482, 492-495 (1977) (grand jury selection). We have in any event recognized that the Free Exercise Clause protects values distinct from those protected by the Equal Protection Clause. See Hobbie, 480 U. S., at 141-142. As the language of the *902 Clause itself makes clear, an individual’s free exercise of religion is a preferred constitutional activity. See, e. g., McConnell, Accommodation of Religion, 1985 S. Ct. Rev. 1, 9 (“[T]he text of the First Amendment itself ‘singles out’ religion for special protections”); P. Kauper, Religion and the Constitution 17 (1964). A law that makes criminal such an activity therefore triggers constitutional concern — and heightened judicial scrutiny — even if it does not target the particular religious conduct at issue. Our free speech cases similarly recognize that neutral regulations that affect free speech values are subject to a balancing, rather than categorical, approach. See, e. g., United States v. O’Brien, 391 U. S. 367, 377 (1968); Renton v. Playtime Theatres, Inc., 475 U. S. 41, 46-47 (1986); cf. Anderson v. Celebrezze, 460 U. S. 780, 792-794 (1983) (generally applicable laws may impinge on free association concerns). The Court’s parade of horribles, ante, at 888-889, not only fails as a reason for discarding the compelling interest test, it instead demonstrates just the opposite: that courts have been quite capable of applying our free exercise jurisprudence to strike sensible balances between religious liberty and competing state interests. Finally, the Court today suggests that the disfavoring of minority religions is an “unavoidable consequence” under our system of government and that accommodation of such religions must be left to the political process. Ante, at 890. In my view, however, the First Amendment was enacted precisely to protect the rights of those whose religious practices are not shared by the majority and may be viewed with hostility. The history of our free exercise doctrine amply demonstrates the harsh impact majoritarian rule has had on unpopular or emerging religious groups such as the Jehovah’s Witnesses and the Amish. Indeed, the words of Justice Jackson in West Virginia State Bd. of Ed. v. Barnette (overruling Minersville School Dist. v. Gobitis, 310 U. S. 586 (1940)) are apt: *903 “The very purpose of a Bill of Rights was to withdraw certain subjects from the vicissitudes of political controversy, to place them beyond the reach of majorities and officials and to establish them as legal principles to be applied by the courts. One’s right to life, liberty, and property, to free speech, a free press, freedom of worship and assembly, and other fundamental rights may not be submitted to vote; they depend on the outcome of no elections.” 319 U. S., at 638. See also United States v. Ballard, 322 U. S. 78, 87 (1944) (“The Fathers of the Constitution were not unaware of the varied and extreme views of religious sects, of the violence of disagreement among them, and of the lack of any one religious creed on which all men would agree. They fashioned a charter of government which envisaged the widest possible toleration of conflicting views”). The compelling interest test reflects the First Amendment’s mandate of preserving religious liberty to the fullest extent possible in a pluralistic society. For the Court to deem this command a “luxury,” ante, at 888, is to denigrate “[t]he very purpose of a Bill of Rights.” Ill The Court’s holding today not only misreads settled First Amendment precedent; it appears to be unnecessary to this case. I would reach the same result applying our established free exercise jurisprudence. A There is no dispute that Oregon’s criminal prohibition of peyote places a severe burden on the ability of respondents to freely exercise their religion. Peyote is a sacrament of the Native American Church and is regarded as vital to respondents’ ability to practice their religion. See O. Stewart, Peyote Religion: A History 327-336 (1987) (describing modern status of peyotism); E. Anderson, Peyote: The Divine Cactus 41-65 (1980) (describing peyote ceremonies); Teachings from *904 the American Earth: Indian Religion and Philosophy 96-104 (D. Tedlock & B. Tedlock eds. 1975) (same); see also People v. Woody, 61 Cal. 2d 716, 721-722, 394 P. 2d 813, 817-818 (1964). As we noted in Smith I, the Oregon Supreme Court concluded that “the Native American Church is a recognized religion, that peyote is a sacrament of that church, and that respondent’s beliefs were sincerely held.” 485 U. S., at 667. Under Oregon law, as construed by that State’s highest court, members of the Native American Church must choose between carrying out the ritual embodying their religious beliefs and avoidance of criminal prosecution. That choice is, in my view, more than sufficient to trigger First Amendment scrutiny. There is also no dispute that Oregon has a significant interest in enforcing laws that control the possession and use of controlled substances by its citizens. See, e. g., Sherbert, 374 U. S., at 403 (religiously motivated conduct may be regulated where such conduct “pose[s] some substantial threat to public safety, peace or order”); Yoder, 406 U. S., at 220 (“[Activities of individuals, even when religiously based, are often subject to regulation by the States in the exercise of their undoubted power to promote the health, safety, and general welfare”). As we recently noted, drug abuse is “one of the greatest problems affecting the health and welfare of our population” and thus “one of the most serious problems confronting our society today.” Treasury Employees v. Von Raab, 489 U. S. 656, 668, 674 (1989). Indeed, under federal law (incorporated by Oregon law in relevant part, see Ore. Rev. Stat. §475.005(6) (1987)), peyote is specifically regulated as a Schedule I controlled substance, which means that Congress has found that it has a high potential for abuse, that there is no currently accepted medical use, and that there is a lack of accepted safety for use of the drug under medical supervision. See 21 U. S. C. § 812(b)(1). See generally R. Julien, A Primer of Drug Action 149 (3d ed. 1981). In light of our recent decisions holding that the governmental *905 interests in the collection of income tax, Hernandez, 490 U. S., at 699-700, a comprehensive Social Security system, see Lee, 455 U. S., at 258-259, and military conscription, see Gillette, 401 U. S., at 460, are compelling, respondents do not seriously dispute that Oregon has a compelling interest in prohibiting the possession of peyote by its citizens. B Thus, the critical question in this case is whether exempting respondents from the State’s general criminal prohibition “will unduly interfere with fulfillment of the governmental interest.” Lee, supra, at 259; see also Roy, 476 U. S., at 727 (“[T]he Government must accommodate a legitimate free exercise claim unless pursuing an especially important interest by narrowly tailored means”); Yoder, supra, at 221; Braunfeld, 366 U. S., at 605-607. Although the question is close, I would conclude that uniform application of Oregon’s criminal prohibition is “essential to accomplish,” Lee, supra, at 257, its overriding interest in preventing the physical harm caused by the use of a Schedule I controlled substance. Oregon’s criminal prohibition represents that State’s judgment that the possession and use of controlled substances, even by only one person, is inherently harmful and dangerous. Because the health effects caused by the use of controlled substances exist regardless of the motivation of the user, the use of such substances, even for religious purposes, violates the very purpose of the laws that prohibit them. Cf. State v. Massey, 229 N. C. 734, 51 S. E. 2d 179 (denying religious exemption to municipal ordinance prohibiting handling of poisonous reptiles), appeal dism’d sub nom. Bunn v. North Carolina, 336 U. S. 942 (1949). Moreover, in view of the societal interest in preventing trafficking in controlled substances, uniform application of the criminal prohibition at issue is essential to the effectiveness of Oregon’s stated interest in preventing any possession of peyote. Cf. Jacobson v. *906 Massachusetts, 197 U. S. 11 (1905) (denying exemption from small pox vaccination requirement). For these reasons, I believe that granting a selective exemption in this case would seriously impair Oregon’s compelling interest in prohibiting possession of peyote by its citizens. Under such circumstances, the Free Exercise Clause does not require the State to accommodate respondents’ religiously motivated conduct. See, e. g., Thomas, 450 U. S., at 719. Unlike in Yoder, where we noted that “[t]he record strongly indicates that accommodating the religious objections of the Amish by forgoing one, or at most two, additional years of compulsory education will not impair the physical or mental health of the child, or result in an inability to be self-supporting or to discharge the duties and responsibilities of citizenship, or in any other way materially detract from the welfare of society,” 406 U. S., at 234; see also id., at 238-240 (White, J., concurring), a religious exemption in this case would be incompatible with the State’s interest in controlling use and possession of illegal drugs. Respondents contend that any incompatibility is belied by the fact that the Federal Government and several States provide exemptions for the religious use of peyote, see 21 CFR § 1307.31 (1989); 307 Ore., at 73, n. 2, 763 P. 2d, at 148, n. 2 (citing 11 state statutes that expressly exempt sacramental peyote use from criminal proscription). But other governments may surely choose to grant an exemption without Oregon, with its specific asserted interest in uniform application of its drug laws, being required to do so by the First Amendment. Respondents also note that the sacramental use of peyote is central to the tenets of the Native American Church, but I agree with the Court, ante, at 886-887, that because “ ‘[i]t is not within the judicial ken to question the centrality of particular beliefs or practices to a faith,’” quoting Hernandez, supra, at 699, our determination of the constitutionality of Oregon’s general criminal prohibition cannot, and should not, turn on the centrality of the particular *907 religious practice at issue. This does not mean, of course, that courts may not make factual findings as to whether a claimant holds a sincerely held religious belief that conflicts with, and thus is burdened by, the challenged law. The distinction between questions of centrality and questions of sincerity and burden is admittedly fine, but it is one that is an established part of our free exercise doctrine, see Ballard, 322 U. S., at 85-88, and one that courts are capable of making. See Tony and Susan Alamo Foundation v. Secretary of Labor, 471 U. S. 290, 303-305 (1985). I would therefore adhere to our established free exercise jurisprudence and hold that the State in this case has a compelling interest in regulating peyote use by its citizens and that accommodating respondents’ religiously motivated conduct “will unduly interfere with fulfillment of the governmental interest.” Lee, supra, at 259. Accordingly, I concur in the judgment of the Court. Justice Blackmun, with whom Justice Brennan and Justice Marshall join, dissenting. This Court over the years painstakingly has developed a consistent and exacting standard to test the constitutionality of a state statute that burdens the free exercise of religion. Such a statute may stand only if the law in general, and the State’s refusal to allow a religious exemption in particular, are justified by a compelling interest that cannot be served by less restrictive means. 1 *908 Until today, I thought this was a settled and inviolate principle of this Court’s First Amendment jurisprudence. The majority, however, perfunctorily dismisses it as a “constitutional anomaly.” Ante, at 886. As carefully detailed in Justice O’Connor’s concurring opinion, ante, p. 891, the majority is able to arrive at this view only by mischaracterizing this Court’s precedents. The Court discards leading free exercise cases such as Cantwell v. Connecticut, 310 U. S. 296 (1940), and Wisconsin v. Yoder, 406 U. S. 205 (1972), as “hybrid.” Ante, at 882. The Court views traditional free exercise analysis as somehow inapplicable to criminal prohibitions (as opposed to conditions on the receipt of benefits), and to state laws of general applicability (as opposed, presumably, to laws that expressly single out religious practices). Ante, at 884-885. The Court cites cases in which, due to various exceptional circumstances, we found strict scrutiny inapposite, to hint that the Court has repudiated that standard altogether. Ante, at 882-884. In short, it effectuates a wholesale overturning of settled law concerning the Religion Clauses of our Constitution. One hopes that the Court is aware of the consequences, and that its result is not a product of overreaction to the serious problems the country’s drug crisis has generated. This distorted view of our precedents leads the majority to conclude that strict scrutiny of a state law burdening the free exercise of religion is a “luxury” that a well-ordered society *909 cannot afford, ante, at 888, and that the repression of minority religions is an “unavoidable consequence of democratic government.” Ante, at 890. I do not believe the Founders thought their dearly bought freedom from religious persecution a “luxury,” but an essential element of liberty — and they could not have thought religious intolerance “unavoidable,” for they drafted the Religion Clauses precisely in order to avoid that intolerance. For these reasons, I agree with Justice O’Connor’s analysis of the applicable free exercise doctrine, and I join parts I and II of her opinion. 2 As she points out, “the critical question in this case is whether exempting respondents from the State’s general criminal prohibition ‘will unduly interfere with fulfillment of the governmental interest.’” Ante, at 905, quoting United States v. Lee, 455 U. S. 252, 259 (1982). I do disagree, however, with her specific answer to that question. I In weighing the clear interest of respondents Smith and Black (hereinafter respondents) in the free exercise of their religion against Oregon’s asserted interest in enforcing its drug laws, it is important to articulate in precise terms the state interest involved. It is not the State’s broad interest *910 in fighting the critical “war on drugs” that must be weighed against respondents’ claim, but the State’s narrow interest in refusing to make an exception for the religious, ceremonial use of peyote. See Bowen v. Roy, 476 U. S. 693, 728 (1986) (O’Connor, J., concurring in part and dissenting in part) (“This Court has consistently asked the Government to demonstrate that unbending application of its regulation to the religious objector ‘is essential to accomplish an overriding governmental interest,’” quoting Lee, 455 U. S., at 257-258); Thomas v. Review Bd. of Indiana Employment Security Div., 450 U. S. 707, 719 (1981) (“focus of the inquiry” concerning State’s asserted interest must be “properly narrowed”); Yoder, 406 U. S., at 221 (“Where fundamental claims of religious freedom are at stake,” the Court will not accept a State’s “sweeping claim” that its interest in compulsory education is compelling; despite the validity of this interest “in the generality of cases, we must searchingly examine the interests that the State seeks to promote . . . and the impediment to those objectives that would flow from recognizing the claimed Amish exemption”). Failure to reduce the competing interests to the same plane of generality tends to distort the weighing process in the State’s favor. See Clark, Guidelines for the Free Exercise Clause, 83 Harv. L. Rev. 327, 330-331 (1969) (“The purpose of almost any law can be traced back to one or another of the fundamental concerns of government: public health and safety, public peace and order, defense, revenue. To measure an individual interest directly against one of these rarified values inevitably makes the individual interest appear the less significant”); Pound, A Survey of Social Interests, 57 Harv. L. Rev. 1, 2 (1943) (“When it comes to weighing or valuing claims or demands with respect to other claims or demands, we must be careful to compare them on the same plane ... [or else] we may decide the question in advance in our very way of putting it”). The State’s interest in enforcing its prohibition, in order to be sufficiently compelling to outweigh a free exercise claim, *911 cannot be merely abstract or symbolic. The State cannot plausibly assert that unbending application of a criminal prohibition is essential to fulfill any compelling interest, if it does not, in fact, attempt to enforce that prohibition. In this case, the State actually has not evinced any concrete interest in enforcing its drug laws against religious users of peyote. Oregon has never sought to prosecute respondents, and does not claim that it has made significant enforcement efforts against other religious users of peyote. 3 The State’s asserted interest thus amounts only to the symbolic preservation of an unenforced prohibition. But a government interest in “symbolism, even symbolism for so worthy a cause as the abolition of unlawful drugs,” Treasury Employees v. Von Raab, 489 U. S. 656, 687 (1989) (Scalia, J., dissenting), cannot suffice to abrogate the constitutional rights of individuals. Similarly, this Court’s prior decisions have not allowed a government to rely on mere speculation about potential harms, but have demanded evidentiary support for a refusal to allow a religious exception. See Thomas, 450 U. S., at 719 (rejecting State’s reasons for refusing religious exemption, for lack of “evidence in the record”); Yoder, 406 U. S., at 224-229 (rejecting State’s argument concerning the dangers of a religious exemption as speculative, and unsupported by the record); Sherbert v. Verner, 374 U. S. 398, 407 (1963) (“[Tjhere is no proof whatever to warrant such fears . . . as those which the [State] now advance[s]”). In this case, the State’s justification for refusing to recognize an exception to its criminal laws for religious peyote use is entirely speculative. The State proclaims an interest in protecting the health and safety of its citizens from the dangers of unlawful drugs. It offers, however, no evidence that the religious use of pey *912 ote has ever harmed anyone. 4 The factual findings of other courts cast doubt on the State’s assumption that religious use of peyote is harmful. See State v. Whittingham, 19 Ariz. App. 27, 30, 504 P. 2d 950, 953 (1973) (“[T]he State failed to prove that the quantities of peyote used in the sacraments of the Native American Church are sufficiently harmful to the health and welfare of the participants so as to permit a legitimate intrusion under the State’s police power”); People v. Woody, 61 Cal. 2d 716, 722-723, 394 P. 2d 813, 818 (1964) (“[A]s the Attorney General . . . admits, . . . the opinion of scientists and other experts is ‘that peyote . . . works no permanent deleterious injury to the Indian’ ”). The fact that peyote is classified as a Schedule I controlled substance does not, by itself, show that any and all uses of peyote, in any circumstance, are inherently harmful and dangerous. The Federal Government, which created the classifications of unlawful drugs from which Oregon’s drug laws are derived, apparently does not find peyote so dangerous as to preclude an exemption for religious use. 5 Moreover, *913 other Schedule I drugs have lawful uses. See Olsen v. Drug Enforcement Admin., 279 U. S. App. D. C. 1, 6, n. 4, 878 F. 2d 1458, 1463, n. 4 (medical and research uses of marijuana). The carefully circumscribed ritual context in which respondents used peyote is far removed from the irresponsible and unrestricted recreational use of unlawful drugs. 6 The Native American Church’s internal restrictions on, and supervision of, its members’ use of peyote substantially obviate the State’s health and safety concerns. See id., at 10, 878 F. 2d, at 1467 (“‘The Administrator [of the Drug Enforcement Administration (DEA)] finds that . . . the Native American Church’s use of peyote is isolated to specific ceremonial occasions,”’ and so “‘an accommodation can be made for a religious organization which uses peyote in circumscribed ceremonies’” (quoting DEA Final Order)); id., at 7, 878 F. 2d, at 1464 (“[F]or members of the Native American Church, use of peyote outside the ritual is sacrilegious”); Woody, 61 Cal. 2d, at 721, 394 P. 2d, at 817 (“[T]o use peyote for nonreligious purposes is sacrilegious”); R. Julien, A Primer of Drug Action 148 (3d ed. 1981) (“[P]eyote is seldom abused by members of the Native American *914 Church”); Slotkin, The Peyote Way, in Teachings from the American Earth 96, 104 (D. Tedlock & B. Tedlock eds. 1975) (“[T]he Native American Church . . . refuses to permit the presence of curiosity seekers at its rites, and vigorously opposes the sale or use of Peyote for non-sacramental purposes”); Bergman, Navajo Peyote Use: Its Apparent Safety, 128 Am. J. Psychiatry 695 (1971) (Bergman). 7 Moreover, just as in Yoder, the values and interests of those seeking a religious exemption in this case are congruent, to a great degree, with those the State seeks to promote through its drug laws. See Yoder, 406 U. S., at 224, 228-229 (since the Amish accept formal schooling up to 8th grade, and then provide “ideal” vocational education, State’s interest in enforcing its law against the Amish is “less substantial than ... for children generally”); id., at 238 (White, J., concurring). Not only does the church’s doctrine forbid nonreligious use of peyote; it also generally advocates self-reliance, familial responsibility, and abstinence from alcohol. See Brief for Association on American Indian Affairs et al. as Amici Curiae 33-34 (the church’s “ethical code” has four parts: brotherly love, care of family, self-reliance, and avoidance of alcohol (quoting from the church membership card)); Olsen, 279 U. S. App. D. C., at 7, 878 F. 2d, at 1464 (the Native American Church, “for all purposes other than the special, stylized ceremony, reinforced the state’s prohibition”); *915 Woody, 61 Cal. 2d, at 721-722, n. 3, 394 P. 2d, at 818, n. 3 (“[M]ost anthropological authorities hold Peyotism to be a positive, rather than negative, force in the lives of its adherents . . . the church forbids the use of alcohol . . .”). There is considerable evidence that the spiritual and social support provided by the church has been effective in combating the tragic effects of alcoholism on the Native American population. Two noted experts on peyotism, Dr. Omer C. Stewart and Dr. Robert Bergman, testified by affidavit to this effect on behalf of respondent Smith before the Employment Appeal Board. Smith Tr., Exh. 7; see also E. Anderson, Peyote: The Divine Cactus 165-166 (1980) (research by Dr. Bergman suggests “that the religious use of peyote seemed to be directed in an ego-strengthening direction with an emphasis on interpersonal relationships where each individual is assured of his own significance as well as the support of the group”; many people have “‘come through difficult crises with the help of this religion .... It provides real help in seeing themselves not as people whose place and way in the world is gone, but as people whose way can be strong enough to change and meet new challenges’ ” (quoting Bergman 698)); Pascarosa & Futterman, Ethnopsychedelic Therapy for Alcoholics: Observations in the Peyote Ritual of the Native American Church, 8 J. of Psychedelic Drugs, No. 3, p. 215 (1976) (religious peyote use has been helpful in overcoming alcoholism); Albaugh & Anderson, Peyote in the Treatment of Alcoholism among American Indians, 131 Am. J. Psychiatry 1247, 1249 (1974) (“[T]he philosophy, teachings, and format of the [Native American Church] can be of great benefit to the Indian alcoholic”); see generally O. Stewart, Peyote Religion 75 et seq. (1987) (noting frequent observations, across many tribes and periods in history, of correlation between peyotist religion and abstinence from alcohol). Far from promoting the lawless and irresponsible use of drugs, Native American Church members’ spiri *916 tual code exemplifies values that Oregon’s drug laws are presumably intended to foster. The State also seeks to support its refusal to make an exception for religious use of peyote by invoking its interest in abolishing drug trafficking. There is, however, practically no illegal traffic in peyote. See Olsen, 279 U. S. App. D. C., at 6, 7, 878 F. 2d, at 1463, 1467 (quoting DEA Final Order to the effect that total amount of peyote seized and analyzed by federal authorities between 1980 and 1987 was 19.4 pounds; in contrast, total amount of marijuana seized during that period was over 15 million pounds). Also, the availability of peyote for religious use, even if Oregon were to allow an exemption from its criminal laws, would still be strictly controlled by federal regulations, see 21 U. S. C. §§821-823 (registration requirements for distribution of controlled substances); 21 CFR § 1307.31 (1989) (distribution of peyote to Native American Church subject to registration requirements), and by the State of Texas, the only State in which peyote grows in significant quantities. See Texas Health & Safety Code Ann. §481.111 (1990 pamphlet); Texas Admin. Code, Tit. 37, pt. 1, ch. 13, Controlled Substances Regulations, §§ 13.35-13.41 (1989); Woody, 61 Cal. 2d, at 720, 394 P. 2d, at 816 (peyote is “found in the Rio Grande Valley of Texas and northern Mexico”). Peyote simply is not a popular drug; its distribution for use in religious rituals has nothing to do with the vast and violent traffic in illegal narcotics that plagues this country. Finally, the State argues that granting an exception for religious peyote use would erode its interest in the uniform, fair, and certain enforcement of its drug laws. The State fears that, if it grants an exemption for religious peyote use, a flood of other claims to religious exemptions will follow. It would then be placed in a dilemma, it says, between allowing a patchwork of exemptions that would hinder its law enforcement efforts, and risking a violation of the Establishment Clause by arbitrarily limiting its religious exemptions. This *917 argument, however, could be made in almost any free exercise case. See Lupu, Where Rights Begin: The Problem of Burdens on the Free Exercise of Religion, 102 Harv. L. Rev. 933, 947 (1989) (“Behind every free exercise claim is a spectral march; grant this one, a voice whispers to each judge, and you will be confronted with an endless chain of exemption demands from religious deviants of every stripe”). This Court, however, consistently has rejected similar arguments in past free exercise cases, and it should do so here as well. See Frazee v. Illinois Dept. of Employment Security, 489 U. S. 829, 835 (1989) (rejecting State’s speculation concerning cumulative effect of many similar claims); Thomas, 450 U. S., at 719 (same); Sherbert, 374 U. S., at 407. The State’s apprehension of a flood of other religious claims is purely speculative. Almost half the States, and the Federal Government, have maintained an exemption for religious peyote use for many years, and apparently have not found themselves overwhelmed by claims to other religious exemptions. 8 Allowing an exemption for religious peyote use *918 would not necessarily oblige the State to grant a similar exemption to other religious groups. The unusual circumstances that make the religious use of peyote compatible with the State’s interests in health and safety and in preventing drug trafficking would not apply to other religious claims. Some religions, for example, might not restrict drug use to a limited ceremonial context, as does the Native American Church. See, e. g., Olsen, 279 U. S. App. D. C., at 7, 878 F. 2d, at 1464 (“[T]he Ethiopian Zion Coptic Church . . . teaches that marijuana is properly smoked ‘continually all day’”). Some religious claims, see n. 8, supra, involve drugs such as marijuana and heroin, in which there is significant illegal traffic, with its attendant greed and violence, so that it would be difficult to grant a religious exemption without seriously compromising law enforcement efforts. 9 That the State might grant an exemption for religious peyote use, but deny other religious claims arising in different circumstances, would not violate the Establishment Clause. Though the State must treat all religions equally, and not favor one over another, this obligation is fulfilled by the uniform application of the “compelling interest” test to all free exercise claims, not by reaching uniform results as to all claims; A showing that religious peyote use does not unduly interfere with the State’s interests is “one that probably few other religious groups or sects could make,” Yoder, 406 U. S., at 236; this does not mean that an exemption limited to peyote use is tantamount to an establishment of religion. See Hobbie v. Unemployment Appeals Comm’n of Fla., 480 U. S. 136, 144-145 (1987) (“[T]he government may (and *919 sometimes must) accommodate religious practices and . . . may do so without violating the Establishment Clause”); Yoder, 406 U. S., at 220-221 (“Court must not ignore the danger that an exception from a general [law] . . . may run afoul of the Establishment Clause, but that danger cannot be allowed to prevent any exception no matter how vital it may be to the protection of values promoted by the right of free exercise”); id., at 234, n. 22. II Finally, although I agree with Justice O’Connor that courts should refrain from delving into questions whether, as a matter of religious doctrine, a particular practice is “central” to the religion, ante, at 906-907, I do not think this means that the courts must turn a blind eye to the severe impact of a State’s restrictions on the adherents of a minority religion. Cf. Yoder, 406 U. S., at 219 (since “education is inseparable from and a part of the basic tenets of their religion . . . [, just as] baptism, the confessional, or a sabbath may be for others,” enforcement of State’s compulsory education law would “gravely endanger if not destroy the free exercise of respondents’ religious beliefs”). Respondents believe, and their sincerity has never been at issue, that the peyote plant embodies their deity, and eating it is an act of worship and communion. Without peyote, they could not enact the essential ritual of their religion. See Brief for Association on American Indian Affairs et al. as Amici Curiae 5-6 (“To the members, peyote is consecrated with powers to heal body, mind and spirit. It is a teacher; it teaches the way to spiritual life through living in harmony and balance with the forces of the Creation. The rituals are an integral part of the life process. They embody a form of worship in which the sacrament Peyote is the means for communicating with the Great Spirit”). See also O. Stewart, Peyote Religion 327-330 (1987) (description of peyote ritual); *920 T. Hillerman, People of Darkness 153 (1980) (description of Navajo peyote ritual). If Oregon can constitutionally prosecute them for this act of worship, they, like the Amish, may be “forced to migrate to some other and more tolerant region.” Yoder, 406 U. S., at 218. This potentially devastating impact must be viewed in light of the federal policy — reached in reaction to many years of religious persecution and intolerance — of protecting the religious freedom of Native Americans. See American Indian Religious Freedom Act, 92 Stat. 469, 42 U. S. C. § 1996 (1982 ed.) (“[I]t shall be the policy of the United States to protect and preserve for American Indians their inherent right of freedom to believe, express, and exercise the traditional religions . . . , including but not limited to access to sites, use and possession of sacred objects, and the freedom to worship through ceremonials and traditional rites”). 10 Congress recognized that certain substances, such as peyote, “have religious significance because they are sacred, they have power, they heal, they are necessary to the exercise of *921 the rites of the religion, they are necessary to the cultural integrity of the tribe, and, therefore, religious survival.” H. R. Rep. No. 95-1308, p. 2 (1978). The American Indian Religious Freedom Act, in itself, may not create rights enforceable against government action restricting religious freedom, but this Court must scrupulously apply its free exercise analysis to the religious claims of Native Americans, however unorthodox they may be. Otherwise, both the First Amendment and the stated policy of Congress will offer to Native Americans merely an unfulfilled and hollow promise. Ill For these reasons, I conclude that Oregon’s interest in enforcing its drug laws against religious use of peyote is not sufficiently compelling to outweigh respondents’ right to the free exercise of their religion. Since the State could not constitutionally enforce its criminal prohibition against respondents, the interests underlying the State’s drug laws cannot justify its denial of unemployment benefits. Absent such justification, the State’s regulatory interest in denying benefits for religiously motivated “misconduct,” see ante, at 874, is indistinguishable from the state interests this Court has rejected in Frazee, Hobbie, Thomas, and Sherbert. The State of Oregon cannot, consistently with the Free Exercise Clause, deny respondents unemployment benefits. I dissent. Footnotes [1] Both lines of cases have specifically adverted to the non-free-exercise principle involved. Cantwell, for example, observed that “[t]he fundamental law declares the interest of the United States that the free exercise of religion be not prohibited and that freedom to communicate information and opinion be not abridged.” 310 U. S., at 307. Murdock said: “We do not mean to say that religious groups and the press are free from all financial burdens of government. . . . We have here something quite different, for example, from a tax on the income of one who engages in religious activities or a tax on property used or employed in connection with those activities. It is one thing to impose a tax on the income or property of a preacher. It is quite another thing to exact a tax from him for the privilege of delivering a sermon. . . . Those who can deprive religious groups of their colporteurs can take from them a part of the vital power of the press which has survived from the Reformation.” 319 U. S., at 112. Yoder said that “the Court’s holding in Pierce stands as a charter of the rights of parents to direct the religious upbringing of their children. And, when the interests of parenthood are combined with a free exercise claim of the nature revealed by this record, more than merely a ‘reasonable relation to some purpose within the competency of the State’ is required to sustain the validity of the State’s requirement under the First Amendment.” 406 U. S., at 233. [2] Justice O’Connor seeks to distinguish Lyng v. Northwest Indian Cemetery Protective Assn., 485 U. S. 439 (1988), and Bowen v. Roy, 476 U. S. 693 (1986), on the ground that those cases involved the government's conduct of “its own internal affairs,” which is different because, as Justice Douglas said in Sherbert, “ ‘the Free Exercise Clause is written in terms of what the government cannot do to the individual, not in terms of what the individual can exact from the government.’ ” Post, at 900 (O’Connor, J., concurring in judgment), quoting Sherbert v. Verner, 374 U. S. 398, 412 (1963) (Douglas, J., concurring). But since Justice Douglas voted with the majority in Sherbert, that quote obviously envisioned that what “the government cannot do to the individual” includes not just the prohibition of an individual’s freedom of action through criminal laws but also the running of its programs (in Sherbert, state unemployment compensation) in such fashion as to harm the individual’s religious interests. Moreover, it is hard to see any reason in principle or practicality why the government should have to tailor its health and safety laws to conform to the diversity of religious belief, but should not have to tailor its management of public lands, Lyng, supra, or its administration of welfare programs, Roy, supra. [3] Justice O’Connor suggests that “[t]here is nothing talismanic about neutral laws of general applicability,” and that all laws burdening religious practices should be subject to compelling-interest scrutiny because “the First Amendment unequivocally makes freedom of religion, like freedom from race discrimination and freedom of speech, a ‘constitutional nor[m],’ not an ‘anomaly.’” Post, at 901 (opinion concurring in judgment). But this comparison with other fields supports, rather than undermines, the conclusion we draw today. Just as we subject to the most exacting scrutiny laws that make classifications based on race, see Palmore v. Sidoti, 466 U. S. 429 (1984), or on the content of speech, see Sable Communications of California v. FCC, 492 U. S. 115 (1989), so too we strictly scrutinize governmental classifications based on religion, see McDaniel v. Paty, 435 U. S. 618 (1978); see also Torcaso v. Watkins, 367 U. S. 488 (1961). But we have held that race-neutral laws that have the effect of disproportionately disadvantaging a particular racial group do not thereby become subject to compelling-interest analysis under the Equal Protection Clause, see Washington v. Davis, 426 U. S. 229 (1976) (police employment examination); and we have held that generally applicable laws unconcerned with regulating speech that have the effect of interfering with speech do not thereby become subject to compelling-interest analysis under the First Amendment, see Citizen Publishing Co. v. United States, 394 U. S. 131, 139 (1969) (antitrust laws). Our conclusion that generally applicable, religion-neutral laws that have the effect of burdening a particular religious practice need not be justified by a compelling governmental interest is the only approach compatible with these precedents. [4] While arguing that we should apply the compelling interest test in this case, Justice O’Connor nonetheless agrees that “our determination of the constitutionality of Oregon’s general criminal prohibition cannot,- and should not, turn on the centrality of the particular religious practice at issue,” post, at 906-907 (opinion concurring in judgment). This means, presumably, that compelling-interest scrutiny must be applied to generally applicable laws that regulate or prohibit any religiously motivated activity, no matter how unimportant to the claimant’s religion. Earlier in her opinion, however, Justice O’Connor appears to contradict this, saying that the proper approach is “to determine whether the burden on the specific plaintiffs before us is constitutionally significant and whether the particular criminal interest asserted by the State before us is compelling.” Post, at 899. “Constitutionally significant burden” would seem to be “cen *888 trality” under another name. In any case, dispensing with a “centrality” inquiry is utterly unworkable. It would require, for example, the same degree of “compelling state interest” to impede the practice of throwing rice at church weddings as to impede the practice of getting married in church. There is no way out of the difficulty that, if general laws are to be subjected to a “religious practice” exception, both the importance of the law at issue and the centrality of the practice at issue must reasonably be considered. Nor is this difficulty avoided by Justice Blackmun’s assertion that “although . . . courts should refrain from delving into questions whether,, as a matter of religious doctrine, a particular practice is ‘central’ to the religion, ... I do not think this means that the courts must turn a blind eye to the severe impact of a State’s restrictions on the adherents of a minority religion.” Post, at 919 (dissenting opinion). As Justice . Blackmun’s opinion proceeds to make clear, inquiry into “severe impact” is no different from inquiry into centrality. He has merely substituted for the question “How important is X to the religious adherent?” the question “How great will be the harm to the religious adherent if X is taken away?” There is no material difference. [5] Justice O’Connor contends that the “parade of horribles” in the text only “demonstrates . . . that courts have been quite capable of . . . striking] sensible balances between religious liberty and competing state interests.” Post, at 902 (opinion concurring in judgment). But the cases we cite have struck “sensible balances” only because they have all applied the general laws, despite the claims for religious exemption. In any event, Justice O’Connor mistakes the purpose of our parade: it is not to suggest that courts would necessarily permit harmful exemptions from these laws (though they might), but to suggest that courts would constantly be in the business of determining whether the “severe impact” of various laws on religious practice (to use Justice Blackmun’s terminology, post, at 919) or the “constitutional!] significance]” of the “burden on the specific plaintiffs” (to use Justice O’Connor’s terminology, post, at 899) suffices to permit us to confer an exemption. It is a parade of horribles because it is horrible to *890 contemplate that federal judges will regularly balance against the importance of general laws the significance of religious practice. [1] See Hernandez v. Commissioner, 490 U. S. 680, 699 (1989) (“The free exercise inquiry asks whether government has placed a substantial burden on the observation of a central religious belief or practice and, if so, whether a compelling governmental interest justifies the burden”); Hobbie v. Unemployment Appeals Comm’n of Fla., 480 U. S. 136, 141 (1987) (state laws burdening religions “must be subjected to strict scrutiny and could be justified only by proof by the State of a compelling interest”); Bowen v. Roy, 476 U. S. 693, 732 (1986) (O’Connor, J., concurring in part and dissenting in part) (“Our precedents have long required the Government to show that a compelling state interest is served by its refusal to grant a religious exemption”); United States v. Lee, 455 U. S. 252, 257-258 *908 (1982) (“The state may justify a limitation on religious liberty by showing that it is essential to accomplish an overriding governmental interest”); Thomas v. Review Bd. of Indiana Employment Security Div., 450 U. S. 707, 718 (1981) (“The state may justify an inroad on religious liberty by showing that it is the least restrictive means of achieving some compelling state interest”); Wisconsin v. Yoder, 406 U. S. 205, 215 (1972) (“[0]nly those interests of the highest order and those not otherwise served can overbalance legitimate claims to the free exercise of religion”); Sherbert v. Verner, 374 U. S. 398, 406 (1963) (question is “whether some compelling state interest. . . justifies the substantial infringement of appellant’s First Amendment right”). [2] 1 reluctantly agree that, in light of this Court’s decision in Employment Division, Dept. of Human Resources of Ore. v. Smith, 485 U. S. 660 (1988), the question on which certiorari was granted is properly presented in this case. I have grave doubts, however, as to the wisdom or propriety of deciding the constitutionality of a criminal prohibition which the State has not sought to enforce, which the State did not rely on in defending its denial of unemployment benefits before the state courts, and which the Oregon courts could, on remand, either invalidate on state constitutional grounds, or conclude that it remains irrelevant to Oregon’s interest in administering its unemployment benefits program. It is surprising, to say the least, that this Court which so often prides itself about principles of judicial restraint and reduction of federal control over matters of state law would stretch its jurisdiction to the limit in order to reach, in this abstract setting, the constitutionality of Oregon’s criminal prohibition of peyote use. [3] The only reported case in which the State of Oregon has sought to prosecute a person for religious peyote use is State v. Soto, 21 Ore. App. 794, 537 P. 2d 142 (1975), cert. denied, 424 U. S. 955 (1976). [4] This dearth of evidence is not surprising, since the State never asserted this health and safety interest before the Oregon courts; thus, there was no opportunity for factfinding concerning the alleged dangers of peyote use. What has now become the State’s principal argument for its view that the criminal prohibition is enforceable against religious use of peyote rests on no evidentiary foundation at all. [5] See 21 CFR § 1307.31 (1989) (“The listing of peyote as a controlled substance in Schedule I does not apply to the nondrug use of peyote in bona fide religious ceremonies of the Native American Church, and members of the Native American Church so using peyote are exempt from registration. Any person who manufactures peyote for or distributes peyote to the Native American Church, however, is required to obtain registration annually and to comply with all other requirements of law”); see Olsen v. Drug Enforcement Admin., 279 U. S. App. D. C. 1, 6-7, 878 F. 2d 1458, 1463-1464 (1989) (explaining DEA’s rationale for the exception). Moreover, 23 States, including many that have significant Native American populations, have statutory or judicially crafted exemptions in their drug laws for religious use of peyote. See 307 Ore. 68, 73, n. 2, 763 P. 2d 146, 148, n. 2 (1988) (case below). Although this does not prove that Ore *913 gon must have such an exception too, it is significant that these States, and the Federal Government, all find their (presumably compelling) interests in controlling the use of dangerous drugs compatible with an exemption for religious use of peyote. Cf. Boos v. Barry, 485 U. S. 312, 329 (1988) (finding that an ordinance restricting picketing near a foreign embassy was not the least restrictive means of serving the asserted government interest; existence of an analogous, but more narrowly drawn, federal statute showed that “a less restrictive alternative is readily available”). [6] In this respect, respondents’ use of peyote seems closely analogous to the sacramental use of wine by the Roman Catholic Church. During Prohibition, the Federal Government exempted such use of wine from its general ban on possession and use of alcohol. See National Prohibition Act, Title II, § 3, 41 Stat. 308. However compelling the Government’s then general interest in prohibiting the use of alcohol may have been, it could not plausibly have asserted an interest sufficiently compelling to outweigh Catholics’ right to take communion. [7] The use of peyote is, to some degree, self-limiting. The peyote plant is extremely bitter, and eating it is an unpleasant experience, which would tend to discourage casual or recreational use. See State v. Whittingham, 19 Ariz. App. 27, 30, 504 P. 2d 950, 953 (1973) (“ ‘[PJeyote can cause vomiting by reason of its bitter taste’”); E. Anderson, Peyote: The Divine Cactus 161 (1980) (“[T]he eating of peyote usually is a difficult ordeal in that nausea and other unpleasant physical manifestations occur regularly. Repeated use is likely, therefore, only if one is a serious researcher or is devoutly involved in taking peyote as part of a religious ceremony”); Slotkin, The Peyote Way, in Teachings from the American Earth 96, 98 (D. Tedlock & B. Tedlock eds. 1975) (“[M]any find it bitter, inducing indigestion or nausea"). [8] Over the years, various sects have raised free exercise claims regarding drug use. In no reported case, except those involving claims of religious peyote use, has the claimant prevailed. See, e. g., Olsen v. Iowa, 808 F. 2d 652 (CA8 1986) (marijuana use by Ethiopian Zion Coptic Church); United States v. Rush, 738 F. 2d 497 (CA1 1984) (same), cert. denied, 470 U. S. 1004 (1985); United States v. Middleton, 690 F. 2d 820 (CA11 1982) (same), cert denied, 460 U. S. 1051 (1983); United States v. Hudson, 431 F. 2d 468 (CA5 1970) (marijuana and heroin use by Moslems), cert denied, 400 U. S. 1011 (1971); Leary v. United States, 383 F. 2d 851 (CA5 1967) (marijuana use by Hindu), rev’d on other grounds, 395 U. S. 6 (1969); Commonwealth v. Nissenbaum, 404 Mass. 575, 536 N. E. 2d 592 (1989) (marijuana use by Ethiopian Zion Coptic Church); State v. Blake, 5 Haw. App. 411, 695 P. 2d 336 (1985) (marijuana use in practice of Hindu Tantrism); Whyte v. United States, 471 A. 2d 1018 (D. C. App. 1984) (marijuana use by Rastafarian); State v. Rocheleau, 142 Vt. 61, 451 A. 2d 1144 (1982) (marijuana use by Tantric Buddhist); State v. Brashear, 92 N. M. 622, 593 P. 2d 63 (1979) (marijuana use by nondenominational Christian); State v. Randall, 540 S. W. 2d 156 (Mo. App. 1976) (marijuana, LSD, and hashish use by Aquarian Brotherhood Church). See generally Annotation, Free *918 Exercise of Religion as Defense to Prosecution for Narcotic or Psychedelic Drug Offense, 35 A. L. R. 3d 939 (1971 and Supp. 1989). [9] Thus, this ease is distinguishable from United States v. Lee, 455 U. S. 252 (1982), in which the Court concluded that there was “no principled way” to distinguish other exemption claims, and the “tax system could not function if denominations were allowed to challenge the tax system because tax payments were spent in a manner that violates their religious belief.” Id., at 260. [10] See Federal Agencies Task Force, Report to Congress on American Indian Religious Freedom Act of 1978, pp. 1-8 (Aug. 1979) (history of religious persecution); Barsh, The Illusion of Religious Freedom for Indigenous Americans, 65 Ore. L. Rev. 363, 369-374 (1986). Indeed, Oregon’s attitude toward respondents’ religious peyote use harkens back to the repressive federal policies pursued a century ago: “In the government’s view, traditional practices were not only morally degrading, but unhealthy. ‘Indians are fond of gatherings of every description,’ a 1913 public health study complained, advocating the restriction of dances and ‘sings’ to stem contagious diseases. In 1921, Commissioner of Indian Affairs Charles Burke reminded his staff to punish any Indian engaged in ‘any dance which involves . . . the reckless giving away of property . . . frequent or prolonged periods of celebration ... in fact, any disorderly or plainly excessive performance that promotes superstitious cruelty, licentiousness, idleness, danger to health, and shiftless indifference to family welfare.’ Two years later, he forbid Indians under the age of 50 from participating in any dances of any kind, and directed federal employees ‘to educate public opinion’ against them.” Id., at 370-371 (footnotes omitted).