[511]*511Justice Kennedy
delivered the opinion of the Court.
A decision by local zoning authorities to deny a church a building permit was challenged under the Religious Freedom Restoration Act of 1993 (RFRA or Act), 107 Stat. 1488, 42 U. S. C. § 2000bb et seq. The case calls into question the authority of Congress to enact RFRA. We conclude the statute exceeds Congress’ power.
I
Situated on a hill in the city of Boerne, Texas, some 28 miles northwest of San Antonio, is St. Peter Catholic Church. Built in 1923, the church’s structure replicates the mission [512]*512style of the region’s earlier history. The church seats about 230 worshippers, a number too small for its growing parish. Some 40 to 60 parishioners cannot be accommodated at some Sunday masses. In order to meet the needs of the congregation the Archbishop of San Antonio gave permission to the parish to plan alterations to enlarge the building.
A few months later, the Boerne City Council passed an ordinance authorizing the city’s Historic Landmark Commission to prepare a preservation plan with proposed historic landmarks and districts. Under the ordinance, the commission must preapprove construction affecting historic landmarks or buildings in a historic district.
Soon afterwards, the Archbishop applied for a building permit so construction to enlarge the church could proceed. City authorities, relying on the ordinance and the designation of a historic district (which, they argued, included the church), denied the application. The Archbishop brought this suit challenging the permit denial in the United States District Court for the Western District of Texas. 877 F. Supp. 355 (1995).
The complaint contained various claims, but to this point the litigation has centered on RFRA and the question of its constitutionality. The Archbishop relied upon RFRA as one basis for relief from the refusal to issue the permit. The District Court concluded that by enacting RFRA Congress exceeded the scope of its enforcement power under § 5 of the Fourteenth Amendment. The court certified its order for interlocutory appeal and the Fifth Circuit reversed, finding RFRA to be constitutional. 73 F. 3d 1352 (1996). We granted certiorari, 519 U. S. 926 (1996), and now reverse.
II
Congress enacted RFRA in direct response to the Court s decision in Employment Div., Dept. of Human Resources of Ore. v. Smith, 494 U. S. 872 (1990). There we considered a Free Exercise Clause claim brought by members of the [513]*513Native American Church who were denied unemployment benefits when they lost their jobs because they had used peyote. Their practice was to ingest peyote for sacramental purposes, and they challenged an Oregon statute of general applicability which made use of the drug criminal. In evaluating the claim, we declined to apply the balancing test set forth in Sherbert v. Verner, 374 U. S. 398 (1963), under which we would have asked whether Oregon’s prohibition substantially burdened a religious practice and, if it did, whether the burden was justified by a compelling government interest. We stated:
“[Gjovernment’s ability to enforce generally applicable prohibitions of socially harmful conduct . . . cannot depend on measuring the effects of a governmental action on a religious objector’s spiritual development. 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’ ... contradicts both constitutional tradition and common sense.” 494 U. S., at 885 (internal quotation marks and citations omitted).
The application of the Sherbert test, the Smith decision explained, would have produced an anomaly in the law, a constitutional right to ignore neutral laws of general applicability. The anomaly would have been accentuated, the Court reasoned, by the difficulty of determining whether a particular practice was central to an individual’s religion. We explained, moreover, that it “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.” 494 U. S., at 887 (internal quotation marks and citation omitted).
The only instances where a neutral, generally applicable law had failed to pass constitutional muster, the Smith Court [514]*514noted, were cases in which other constitutional protections were at stake. Id., at 881-882. In Wisconsin v. Yoder, 406 U. S. 205 (1972), for example, we invalidated Wisconsin’s mandatory school-attendance law as applied to Amish parents who refused on religious grounds to send their children to school. That case implicated not only the right to the free exercise of religion but also the right of parents to control their children’s education.
The Smith decision acknowledged the Court had employed the Sherbert test in considering free exercise challenges to state unemployment compensation rules on three occasions where the balance had tipped in favor of the individual. See Sherbert, supra; Thomas v. Review Bd. of Indiana Employment Security Div., 450 U. S. 707 (1981); Hobbie v. Unemployment Appeals Comm’n of Fla., 480 U. S. 136 (1987). Those cases, the Court explained, 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.” 494 U. S., at 884 (internal quotation marks omitted). By contrast, where a general prohibition, such as Oregon’s, is at issue, “the sounder approach, and the approach in accord with the vast majority of our precedents, is to hold the test inapplicable to [free exercise] challenges.” Id., at 885. Smith held that neutral, generally applicable laws may be applied to religious practices even when not supported by a compelling governmental interest.
Four Members of the Court disagreed. They argued the law placed a substantial burden on the Native American Church members so that it could be upheld only if the law served a compelling state interest and was narrowly tailored to achieve that end. Id., at 894. Justice O’Connor concluded Oregon had satisfied the test, while Justice Blackmun, joined by Justice Brennan and Justice Marshall, could see no compelling interest justifying the law’s application to the members.
[515]*515These points of constitutional interpretation were debated by Members of Congress in hearings and floor debates. Many criticized the Court’s reasoning, and this disagreement resulted in the passage of RFRA. Congress announced:
“(1) [T]he framers of the Constitution, recognizing free exercise of religion as an unalienable right, secured its protection in the First Amendment to the Constitution; “(2) laws 'neutral’ toward religion may burden religious exercise as surely as laws intended to interfere with religious exercise;
“(3) governments should not substantially burden religious exercise without compelling justification;
“(4) in Employment Division v. Smith, 494 U. S. 872 (1990), the Supreme Court virtually eliminated the requirement that the government justify burdens on religious exercise imposed by laws neutral toward religion; and
“(5) the compelling interest test as set forth in prior Federal court rulings is a workable test for striking sensible balances between religious liberty and competing prior governmental interests.” 42 U. S. C. § 2000bb(a).
The Act’s stated purposes are:
“(1) to restore the compelling interest test as set forth in Sherbert v. Verner, 374 U. S. 398 (1963) and Wisconsin v. Yoder, 406 U. S. 205 (1972) and to guarantee its application in all cases where free exercise of religion is substantially burdened; and
“(2) to provide a claim or defense to persons whose religious exercise is substantially burdened by government.” §2000bb(b).
RFRA prohibits “[government” from “substantially bur-denting]” a person’s exercise of religion even if the burden results from a rule of general applicability unless the government can demonstrate the burden “(1) is in furtherance of [516]*516a compelling governmental interest; and (2) is the least restrictive means of furthering that compelling governmental interest.” §2000bb-l. The Act’s mandate applies to any “branch, department, agency, instrumentality, and official (or other person acting under color of law) of the United States,” as well as to any “State, or . . . subdivision of a State.” § 2000bb-2(l). ' The Act’s universal coverage is confirmed in § 2000bb-3(a), under which RFRA “applies to all Federal and State law, and the implementation of that law, whether statutory or otherwise, and whether adopted before or after [RFRA’s enactment].” In accordance with RFRA’s usage of the term, we shall use “state law” to include local and municipal ordinances.
III
A
Under our Constitution, the Federal Government is one of enumerated powers. McCulloch v. Maryland, 4 Wheat. 316, 405 (1819); see also The Federalist No. 45, p. 292 (C. Rossiter ed. 1961) (J. Madison). The judicial authority to determine the constitutionality of laws, in cases and controversies, is based on the premise that the “powers of the legislature are defined and limited; and that those limits may not be mistaken, or forgotten, the constitution is written.” Marbury v. Madison, 1 Cranch 137, 176 (1803).
Congress relied on its Fourteenth Amendment enforcement power in enacting the most far-reaching and substantial of RFRA’s provisions, those which impose its requirements on the States. See Religious Freedom Restoration Act of 1993, S. Rep. No. 103-111, pp. 13-14 (1993) (Senate Report); H. R. Rep. No. 103-88, p. 9 (1993) (House Report). The Fourteenth Amendment provides, in relevant part:
“Section 1. ... No State shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States; nor shall any State deprive any person of life, liberty, or property, without due proc[517]*517ess of law; nor deny to any person within its jurisdiction the equal protection of the laws.
“Section 5. The Congress shall have power to enforce, by appropriate legislation, the provisions of this article.”
The parties disagree over whether RFRA is a proper exercise of Congress’ §5 power “to enforce” by “appropriate legislation” the constitutional guarantee that no State shall deprive any person of “life, liberty, or property, without due process of law,” nor deny any person “equal protection of the laws.”
In defense of the Act, respondent the Archbishop contends, with support from the United States, that RFRA is permissible enforcement legislation. Congress, it is said, is only protecting by legislation one of the liberties guaranteed by the Fourteenth Amendment’s Due Process Clause, the free exercise of religion, beyond what is necessary under Smith. It is said the congressional decision to dispense with proof of deliberate or overt discrimination and instead concentrate on a law’s effects accords with the settled understanding that § 5 includes the power to enact legislation designed to prevent, as well as remedy, constitutional violations. It is further contended that Congress’ § 5 power is not limited to remedial or preventive legislation.
All must acknowledge that § 5 is “a positive grant of legislative power” to Congress, Katzenbach v. Morgan, 384 U. S. 641, 651 (1966). In Ex parte Virginia, 100 U. S. 339, 345-346 (1880), we explained the scope of Congress’ § 5 power in the following broad terms:
“Whatever legislation is appropriate, that is, adapted to carry out the objects the amendments have in view, whatever tends to enforce submission to the prohibitions they contain, and to secure to all persons the enjoyment of perfect equality of civil rights and the equal protection of the laws against State denial or invasion, if not [518]*518prohibited, is brought within the domain of congressional power.”
Legislation which deters or remedies constitutional violations can fall within the sweep of Congress’ enforcement power even if in the process it prohibits conduct which is not itself unconstitutional and intrudes into “legislative spheres of autonomy previously reserved to the States.” Fitzpatrick v. Bitzer, 427 U. S. 445, 455 (1976). For example, the Court upheld a suspension of literacy tests and similar voting requirements under Congress’ parallel power to enforce the provisions of the Fifteenth Amendment, see U. S. Const., Amdt. 15, §2, as a measure to combat racial discrimination in voting, South Carolina v. Katzenbach, 383 U. S. 301, 308 (1966), despite the facial constitutionality of the tests under Lassiter v. Northampton County Bd. of Elections, 360 U. S. 45 (1959). We have also concluded that other measures protecting voting rights are within Congress’ power to enforce the Fourteenth and Fifteenth Amendments, despite the burdens those measures placed on the States. South Carolina v. Katzenbach, supra (upholding several provisions of the Voting Rights Act of 1965); Katzenbach v. Morgan, supra (upholding ban on literacy tests that prohibited certain people schooled in Puerto Rico from voting); Oregon v. Mitchell, 400 U. S. 112 (1970) (upholding 5-year nationwide ban on literacy tests and similar voting requirements for registering to vote); City of Rome v. United States, 446 U. S. 156, 161 (1980) (upholding 7-year extension of the Voting Rights Act’s requirement that certain jurisdictions preclear any change to a “ ‘standard, practice, or procedure with respect to voting’ ”); see also James Everard’s Breweries v. Day, 265 U. S. 545 (1924) (upholding ban on medical prescription of intoxicating malt liquors as appropriate to enforce Eighteenth Amendment ban on manufacture, sale, or transportation of intoxicating liquors for beverage purposes).
It is also true, however, that “[a]s broad as the congressional enforcement power is, it is not unlimited.” Oregon v. [519]*519Mitchell, supra, at 128 (opinion of Black, J.). In assessing the breadth of § 5’s enforcement power, we begin with its text. Congress has been given the power “to enforce” the “provisions of this article.” We agree with respondent, of course, that Congress can enact legislation under § 5 enforcing the constitutional right to the free exercise of religion. The “provisions of this article,” to which §5 refers, include the Due Process Clause of the Fourteenth Amendment. Congress’ power to enforce the Free Exercise Clause follows from our holding in Cantwell v. Connecticut, 310 U. S. 296, 303 (1940), that the “fundamental concept of liberty embodied in [the Fourteenth Amendment’s Due Process Clause] embraces the liberties guaranteed by the First Amendment.” See also United States v. Price, 383 U. S. 787, 789 (1966) (there is “no doubt of the power of Congress to enforce by appropriate criminal sanction every right guaranteed by the Due Process Clause of the Fourteenth Amendment” (internal quotation marks and citation omitted)).
Congress’ power under § 5, however, extends only to “enforcing]” the provisions of the Fourteenth Amendment. The Court has described this power as “remedial,” South Carolina v. Katzenbach, supra, at 326. The design of the Amendment and the text of § 5 are inconsistent with the suggestion that Congress has the power to decree the substance of the Fourteenth Amendment’s restrictions on the States. Legislation which alters the meaning of the Free Exercise Clause cannot be said to be enforcing the Clause. Congress does not enforce a constitutional right by changing what the right is. It has been given the power “to enforce,” not the power to determine what constitutes a constitutional violation. Were it not so, what Congress would be enforcing would no longer be, in any meaningful sense, the “provisions of [the Fourteenth Amendment].”
While the line between measures that remedy or prevent unconstitutional actions and measures that make a substantive change in the governing law is not easy to discern, and [520]*520Congress must have wide latitude in determining where it lies, the distinction exists and must be observed. There must-be a congruence and proportionality between the injury to be prevented or remedied and the means adopted to that end. Lacking such a connection, legislation may become substantive in operation and effect. History and our case law support drawing the distinction, one apparent from the text of the Amendment.
The Fourteenth Amendment’s history confirms the remedial, rather than substantive, nature of the Enforcement Clause. The Joint Committee on Reconstruction of the 39th Congress began drafting what would become the Fourteenth Amendment in January 1866. The objections to the Committee’s first draft of the Amendment, and the rejection of the draft, have a direct bearing on the central issue of defining Congress’ enforcement power. In February, Republican Representative John Bingham of Ohio reported the following draft Amendment to the House of Representatives on behalf of the Joint Committee:
“The Congress shall have power to make all laws which shall be necessary and proper to secure to the citizens of each State all privileges and immunities of citizens in the several States, and to all persons in the several States equal protection in the rights of life, liberty, and property.” Cong. Globe, 39th Cong., 1st Sess., 1034 (1866).
The proposal encountered immediate opposition, which continued through three days of debate. Members of Congress from across the political spectrum criticized the Amendment, and the criticisms had a common theme: The proposed Amendment gave Congress too much legislative power at the expense of the existing constitutional structure. E. g., id., at 1063-1065 (statement of Rep. Hale); id., at 1082 [521]*521(statement of Sen. Stewart); id., at 1095 (statement of Rep. Hotchkiss); id., at App. 133-135 (statement of Rep. Rogers). Democrats and conservative Republicans argued that the proposed Amendment would give Congress a power to intrude into traditional areas of state responsibility, a power inconsistent with the federal design central to the Constitution. Typifying these views, Republican Representative Robert Hale of New York labeled the Amendment “an utter departure from every principle ever dreamed of by the men who framed our Constitution,” id., at 1063, and warned that under it “all State legislation, in its codes of civil and criminal jurisprudence and procedure . . . may be overridden, may be repealed or abolished, and the law of Congress established instead.” Ibid. Senator William Stewart of Nevada likewise stated the Amendment would permit “Congress to legislate fully upon all subjects affecting life, liberty, and property,” such that “there would not be much left for the State Legislatures,” and would thereby “work an entire change in our form of government.” Id., at 1082; accord, id., at 1087 (statement of Rep. Davis); id., at App. 133 (statement of Rep. Rogers). Some radicals, like their brethren “unwilling that Congress shall have any such power ... to establish uniform laws throughout the United States upon . . . the protection of life, liberty, and property,” id., at 1095 (statement of Rep. Hotchkiss), also objected that giving Congress primary responsibility for enforcing legal equality would place power in the hands of changing congressional majorities, ibid. See generally Bickel, The Original Understanding and the Segregation Decision, 69 Harv. L. Rev. 1, 57 (1955); Graham, Our “Declaratory” Fourteenth Amendment, 7 Stan. L. Rev. 3, 21 (1954).
As a result of these objections having been expressed from so many different quarters, the House voted to table the proposal until April. See, e. g., B. Kendrick, Journal of the Joint Committee of Fifteen on Reconstruction 215, 217 (1914); Cong. Globe, 42d Cong., 1st Sess., App. 115 (1871) (statement [522]*522of Rep. Farnsworth). The congressional action was seen as marking the defeat of the proposal. See The Nation, Mar. 8, 1866, p. 291 (“The postponement of the amendment... is conclusive against the passage of [it]”); New York Times, Mar. 1, 1866, p. 4 (“It is doubtful if this ever comes before the House again . . .”); see also Cong. Globe, 42d Cong., 1st Sess., at App. 115 (statement of Rep. Farnsworth) (The Amendment was “given its quietus by a postponement for two months, where it slept the sleep that knows no waking”). The measure was defeated “chiefly because many members of the legal profession s[aw] in [it]... a dangerous centralization of power,” The Nation, supra, at 291, and “many leading Republicans of th[e] House [of Representatives] would not consent to so radical a change in the Constitution,” Cong. Globe, 42d Cong., 1st Sess., at App. 151 (statement of Rep. Garfield). The Amendment in its early form was not again considered. Instead, the Joint Committee began drafting a new article of Amendment, which it reported to Congress on April 30, 1866.
Section 1 of the new draft Amendment imposed self-executing limits on the States. Section 5 prescribed that “[t]he Congress shall have power to enforce, by appropriate legislation, the provisions of this article.” See Cong. Globe, 39th Cong., 1st Sess., at 2286. Under the revised Amendment, Congress’ power was no longer plenary but remedial. Congress was granted the power to make the substantive constitutional prohibitions against the States effective. Representative Bingham said the new draft would give Congress “the power ... to protect by national law the privileges and immunities of all the citizens of the Republic . . . whenever the same shall be abridged or denied by the unconstitutional acts of any State.” Id., at 2542. Representative Stevens described the new draft Amendment as “allowing] Congress to correct the unjust legislation of the States.” Id., at 2459. See also id., at 2768 (statement of Sen. Howard) (§5 “enables Congress, in case the States shall enact [523]*523laws in conflict with the principles of the amendment, to correct that legislation by a formal congressional enactment”). See generally H. Brannon, The Rights and Privileges Guaranteed by the Fourteenth Amendment to the Constitution of the United States 387 (1901) (Congress’ “powers are only prohibitive, corrective, vetoing, aimed only at undue process of law”); id., at 420, 452-455 (same); T. Cooley, Constitutional Limitations 294, n. 1 (2d ed. 1871) (“This amendment of the Constitution does not concentrate power in the general government for any purpose of police government within the States; its object is to preclude legislation by any State which shall ‘abridge the privileges or immunities of citizens of the United States’ ”). The revised Amendment proposal did not raise the concerns expressed earlier regarding broad congressional power to prescribe uniform national laws with respect to life, liberty, and property. See, e. g., Cong. Globe, 42d Cong., 1st Sess., at App. 151 (statement of Rep. Garfield) (“The [Fourteenth Amendment] limited but did not oust the jurisdiction of the State[s]”). After revisions not relevant here, the new measure passed both Houses and was ratified in July 1868 as the Fourteenth Amendment.
The significance of the defeat of the Bingham proposal was apparent even then. During the debates over the Ku Klux Klan Act only a few years after the Amendment’s ratification, Representative James Garfield argued there were limits on Congress’ enforcement power, saying “unless we ignore both the history and the language of these clauses we cannot, by any reasonable interpretation, give to [§ 5]... the force and effect of the rejected [Bingham] clause.” Ibid.; see also id., at App. 115-116 (statement of Rep. Farnsworth). Scholars of successive generations have agreed with this assessment. See H. Flack, The Adoption of the Fourteenth Amendment 64 (1908); Bickel, The Voting Rights Cases, 1966 S. Ct. Rev. 79, 97.
The design of the Fourteenth Amendment has proved significant also in maintaining the traditional separation of pow[524]*524ers between Congress and the Judiciary. The first eight Amendments to the Constitution set forth self-executing prohibitions on governmental action, and this Court has had primary authority to interpret those prohibitions. The Bing-ham draft, some thought, departed from that tradition by vesting in Congress primary power to interpret and elaborate on the-meaning of the new Amendment through legislation. Under it, “Congress, and not the courts, was to judge whether or not any of the privileges or immunities were not secured to citizens in the several States.” Flack, supra, at 64. While this separation-of-powers aspect did not occasion the widespread resistance which was caused by the proposal’s threat to the federal balance, it nonetheless attracted the attention of various Members. See Cong. Globe, 39th Cong., 1st Sess., at 1064 (statement of Rep. Hale) (noting that Bill of Rights, unlike the Bingham proposal, “provided] safeguards to be enforced by the courts, and not to be exercised by the Legislature”); id., at App. 133 (statement of Rep. Rogers) (prior to Bingham proposal it “was left entirely for the courts ... to enforce the privileges and immunities of the citizens”). As enacted, the Fourteenth Amendment confers substantive rights against the States which, like the provisions of the Bill of Rights, are self-executing. Cf. South Carolina v. Katzenbach, 383 U. S., at 325 (discussing Fifteenth Amendment). The power to interpret the Constitution in a case or controversy remains in the Judiciary.
The remedial and preventive nature of Congress’ enforcement power, and the limitation inherent in the power, were confirmed in our earliest cases on the Fourteenth Amendment. In the Civil Rights Cases, 109 U. S. 3 (1883), the Court invalidated sections of the Civil Rights Act of 1875 which prescribed criminal penalties for denying to any person “the full enjoyment of” public accommodations and conveyances, on the grounds that it exceeded Congress’ power [525]*525by seeking to regulate private conduct. The Enforcement Clause, the Court said, did not authorize Congress to pass “general legislation upon the rights of the citizen, but corrective legislation, that is, such as may be necessary and proper for counteracting such laws as the States may adopt or enforce, and which, by the amendment, they are prohibited from making or enforcing . . . Id., at 13-14. The power to “legislate generally upon” life, liberty, and property, as opposed to the “power to provide modes of redress” against offensive state action, was “repugnant” to the Constitution. Id., at 15. See also United States v. Reese, 92 U. S. 214, 218 (1876); United States v. Harris, 106 U. S. 629, 639 (1883); James v. Bowman, 190 U. S. 127, 139 (1903). Although the specific holdings of these early cases might have been superseded or modified, see, e. g., Heart of Atlanta Motel, Inc. v. United States, 379 U. S. 241 (1964); United States v. Guest, 383 U. S. 745 (1966), their treatment of Congress’ § 5 power as corrective or preventive, not definitional, has not been questioned.
Recent cases have continued to revolve around the question whether §5 legislation can be considered remedial. In South Carolina v. Katzenbach, supra, we emphasized that “[t]he constitutional propriety of [legislation adopted under the Enforcement Clause] must be judged with reference to the historical experience ... it reflects.” 383 U. S., at 308. There we upheld various provisions of the Voting Rights Act of 1965, finding them to be “remedies aimed at areas where voting discrimination has been most flagrant,” id., at 315, and necessary to “banish the blight of racial discrimination in voting, which has infected the electoral process in parts of our country for nearly a century,” id., at 308. We noted evidence in the record reflecting the subsisting and pervasive discriminatory — and therefore unconstitutional — use of literacy tests. Id., at 333-334. The Act’s new remedies, which used the administrative resources of the Federal Government, included the suspension of both literacy tests and, [526]*526pending federal review, all new voting regulations in covered jurisdictions, as well as the assignment of federal examiners to list qualified applicants enabling those listed to vote. The new, unprecedented remedies were deemed necessary given the ineffectiveness of the existing voting rights laws, see id., at 313-315, and the slow, costly character of case-by-case litigation, id., at 328.
After South Carolina v. Katzenbach, the Court continued to acknowledge the necessity of using strong remedial and preventive measures to respond to the widespread and persisting deprivation of constitutional rights resulting from this country’s history of racial discrimination. See Oregon v. Mitchell, 400 U. S., at 132 (“In enacting the literacy test ban . . . Congress had before it a long history of the discriminatory use of literacy tests to disfranchise voters on account of their race”) (opinion of Black, J.); id., at 147 (Literacy tests “have been used at times as a discriminatory weapon against some minorities, not only Negroes but Americans of Mexican ancestry, and American Indians”) (opinion of Douglas, J.); id., at 216 (“Congress could have determined that racial prejudice is prevalent throughout the Nation, and that literacy tests unduly lend themselves to discriminatory application, either conscious or unconscious”) (opinion of Harlan, J.); id., at 235 (“[T]here is no question but that Congress could legitimately have concluded that the use of literacy tests anywhere within the United States has the inevitable effect of denying the vote to members of racial minorities whose inability to pass such tests is the direct consequence of previous governmental discrimination in education”) (opinion of Brennan, J.); id., at 284 (“[Nationwide [suspension of literacy tests] may be reasonably thought appropriate when Congress acts against an evil such as racial discrimination which in varying degrees manifests itself in every part of the country”) (opinion of Stewart, J.); City of Rome, 446 U. S., at 182 (“Congress’ considered determination that at least another 7 years of statutory remedies were necessary to counter the [527]*527perpetuation of 95 years of pervasive voting discrimination is both unsurprising and unassailable”); Morgan, 384 U. S., at 656 (Congress had a factual basis to conclude that New York’s literacy requirement “constituted an invidious discrimination in violation of the Equal Protection Clause”).
Any suggestion that Congress has a substantive, non-remedial power under the Fourteenth Amendment is not supported by our case law. In Oregon v. Mitchell, supra, at 112, a majority of the Court concluded Congress had exceeded its enforcement powers by enacting legislation lowering the minimum age of voters from 21 to 18 in state and local elections. The five Members of the Court who reached this conclusion explained that the legislation intruded into an area reserved by the Constitution to the States. See 400 U. S., at 125 (concluding that the legislation was unconstitutional because the Constitution “reserves to the States the power to set voter qualifications in state and local elections”) (opinion of Black, J.); id., at 154 (explaining that the “Fourteenth Amendment was never intended to restrict the authority of the States to allocate their political power as they see fit”) (opinion of Harlan, J.); id., at 294 (concluding that States, not Congress, have the power “to establish a qualification for voting based on age”) (opinion of Stewart, J., joined by Burger, C. J., and Blackmun, J.). Four of these five were explicit in rejecting the position that § 5 endowed Congress with the power to establish the meaning of constitutional provisions. See id., at 209 (opinion of Harlan, J.); id., at 296 (opinion of Stewart, J.). Justice Black’s rejection of this position might be inferred from his disagreement with Congress’ interpretation of the Equal Protection Clause. See id., at 125.
There is language in our opinion in Katzenbach v. Morgan, 384 U. S. 641 (1966), which could be interpreted as acknowledging a power in Congress to enact legislation that expands [528]*528the rights contained in § 1 of the Fourteenth Amendment. This is not a necessary interpretation, however, or even the best one. In Morgan, the Court considered the constitutionality of §4(e) of the Voting Rights Act of 1965, which provided that no person who had successfully completed the sixth primary grade in a public school in, or a private school accredited by, the Commonwealth of Puerto Rico in which the language of instruction was other than English could be denied the right to vote because of an inability to read or write English. New York’s Constitution, on the other hand, required voters to be able to read and write English. The Court provided two related rationales for its conclusion that § 4(e) could “be viewed as a measure to secure for the Puerto Rican community residing in New York nondiscriminatory treatment by government.” Id., at 652. Under the first rationale, Congress could prohibit New York from denying the right to vote to large segments of its Puerto Rican community, in order to give Puerto Ricans “enhanced political power” that would be “helpful in gaining nondiscriminatory treatment in public services for the entire Puerto Rican community.” Ibid. Section 4(e) thus could be justified as a remedial measure to deal with “discrimination in governmental services.” Id., at 653. The second rationale, an alternative holding, did not address discrimination in the provision of public services but “discrimination in establishing voter qualifications.” Id., at 654. The Court perceived a factual basis on which Congress could have concluded that New York’s literacy requirement “constituted an invidious discrimination in violation of the Equal Protection Clause.” Id., at 656. Both rationales for upholding § 4(e) rested on unconstitutional discrimination by New York and Congress’ reasonable attempt to combat it. As Justice Stewart explained in Oregon v. Mitchell, supra, at 296, interpreting Morgan to give Congress the power to interpret the Constitution “would require an enormous extension of that decision’s rationale.”
[529]*529If Congress could define its own powers by altering the Fourteenth Amendment’s meaning, no longer would the Constitution be “superior paramount law, unchangeable by ordinary means.” It would be “on a level with ordinary legislative acts, and, like other acts, . . . alterable when the legislature shall please to alter it.” Marbury v. Madison, 1 Cranch, at 177. Under this approach, it is difficult to conceive of a principle that would limit congressional power. See Van Alstyne, The Failure of the Religious Freedom Restoration Act under Section 5 of the Fourteenth Amendment, 46 Duke L. J. 291, 292-303 (1996). Shifting legislative majorities could change the Constitution and effectively circumvent the difficult and detailed amendment process contained in Article V.
We now turn to consider whether RFRA can be considered enforcement legislation under §5 of the Fourteenth Amendment.
B
Respondent contends that RFRA is a proper exercise of Congress’ remedial or preventive power. The Act, it is said, is a reasonable means of protecting the free exercise of religion as defined by Smith. It prevents and remedies laws which are enacted with the unconstitutional object of targeting religious beliefs and practices. See Church of Lukumi Babalu Aye, Inc. v. Hialeah, 508 U. S. 520, 533 (1993) (“[A] law targeting religious beliefs as such is never permissible”). To avoid the difficulty of proving such violations, it is said, Congress can simply invalidate any law which imposes a substantial burden on a religious practice unless it is justified by a compelling interest and is the least restrictive means of accomplishing that interest. If Congress can prohibit laws with discriminatory effects in order to prevent racial discrimination in violation of the Equal Protection Clause, see Fullilove v. Klutznick, 448 U. S. 448, 477 (1980) (plurality opinion); City of Rome, 446 U. S., at 177, then it can do the same, respondent argues, to promote religious liberty.
[530]*530While preventive rules are sometimes appropriate remedial measures, there must be a congruence between the means used and the ends to be achieved. The appropriateness of remedial measures must be considered in light of the evil presented. See South Carolina v. Katzenbach, 383 U. S., at 308. Strong measures appropriate to address one harm may be an unwarranted response to another, lesser one. Id., at 334.
A comparison between RFRA and the Voting Rights Act is instructive. In contrast to the record which confronted Congress and the Judiciary in the voting rights cases, RFRA’s legislative record lacks examples of modern instances of generally applicable laws passed because of religious bigotry. The history of persecution in this country detailed in the hearings mentions no episodes occurring in the past 40 years. See, e. g., Religious Freedom Restoration Act of 1991, Hearings on H. R. 2797 before the Subcommittee on Civil and Constitutional Rights of the House Committee on the Judiciary, 102d Cong., 2d Sess., 331-334 (1993) (statement of Douglas Laycock) (House Hearings); The Religious Freedom Restoration Act, Hearing on S. 2969 before the Senate Committee on the Judiciary, 102d Cong., 2d Sess., 30-31 (1993) (statement of Dallin H. Oaks) (Senate Hearing); id., at 68-76 (statement of Douglas Laycock); Religious Freedom Restoration Act of 1990, Hearing on H. R. 5377 before the Subcommittee on Civil and Constitutional Rights of the House Committee on the Judiciary, 101st Cong., 2d Sess., 49 (1991) (statement of John H. Buchanan, Jr.) (1990 House Hearing). The absence of more recent episodes stems from the fact that, as one witness testified, “deliberate persecution is not the usual problem in this country.” House Hearings 334 (statement of Douglas Laycock). See also House Report 2 (“[L]aws directly targeting religious practices have become increasingly rare”). Rather, the emphasis of the hearings was on laws of general applicability which place incidental burdens on religion. Much of the discussion cen[531]*531tered upon anecdotal evidence of autopsies performed on Jewish individuals and Hmong immigrants in violation of their religious beliefs, see, e. g., House Hearings 81 (statement of Nadine Strossen); id., at 107-110 (statement of William Yang); id., at 118 (statement of Rep. Stephen J. Solarz); id., at 336 (statement of Douglas Laycock); Senate Hearing 5-6, 14-26 (statement of William Yang); id., at 27-28 (statement of Hmong-Lao Unity Assn., Inc.); id., at 50 (statement of Baptist Joint Committee); see also Senate Report 8; House Report 5-6, and n. 14, and on zoning regulations and historic preservation laws (like the one at issue here), which, as an incident of their normal operation, have adverse effects on churches and synagogues. See, e. g., House Hearings 17, 57' (statement of Robert P. Dugan, Jr.); id., at 81 (statement of Nadine Strossen); id., at 122-123 (statement of Rep. Stephen J. Solarz); id., at 157 (statement of Edward M. Gaffney, Jr.); id., at 327 (statement of Douglas Laycock); Senate Hearing 143-144 (statement of Forest D. Montgomery); 1990 House Hearing 39 (statement of Robert P. Dugan, Jr.); see also Senate Report 8; House Report 5-6, and n. 14. It is difficult to. main tain that they are examples of legislation enacted or enforced due to animus or hostility to the burdened religious practices or that they indicate some widespread pattern of religious discrimination in this country. Congress’ concern was with the incidental burdens imposed, not the object or purpose of the legislation. See House Report 2; Senate Report 4-5; House Hearings 64 (statement of Nadine Strossen); id., at 117-118 (statement of Rep. Stephen J. Solarz); 1990 House Hearing 14 (statement of Rep. Stephen J. Solarz). This lack of support in the legislative record, however, is not RFRA’s most serious shortcoming. Judicial deference, in most cases, is based not on the state of the legislative record Congress compiles but “on due regard for the decision of the body constitutionally appointed to decide.” Oregon v. Mitchell, 400 U. S., at 207 (opinion of Harlan, J.). As a gen[532]*532eral matter, it is for Congress to determine the method by which it will reach a decision.
Regardless of the state of the legislative record, RFRA cannot be considered remedial, preventive legislation, if those terms are to have any meaning. RFRA is so out of proportion to a supposed remedial or preventive object that it cannot be understood as responsive to, or designed to prevent, unconstitutional behavior. It appears, instead, to attempt a substantive change in constitutional protections. Preventive measures prohibiting certain types of laws may be appropriate when there is reason to believe that many of the laws affected by the congressional enactment have a significant likelihood of being unconstitutional. See City of Rome, 446 U. S., at 177 (since “jurisdictions with a demonstrable history of intentional racial discrimination . . . create the risk of purposeful discrimination,” Congress could “prohibit changes that have a discriminatory impact” in those jurisdictions). Remedial legislation under § 5 “should be adapted to the mischief and wrong which the [Fourteenth] [AJmendment was intended to provide against.” Civil Rights Cases, 109 U. S., at 13.
RFRA is not so confined. Sweeping coverage ensures its intrusion at every level of government, displacing laws and prohibiting official actions of almost every description and regardless of subject matter. RFRA’s restrictions apply to every agency and official of the Federal, State, and local Governments. 42 U. S. C. § 2000bb-2(1). RFRA applies to all federal and state law, statutory or otherwise, whether adopted before or after its enactment. § 2000bb-3(a). RFRA has no termination date or termination mechanism. Any law is subject to challenge at any time by any individual who alleges a substantial burden on his or her free exercise of religion.
The reach and scope of RFRA distinguish it from other measures passed under Congress’ enforcement power, even in the area of voting rights. In South Carolina v. Katzenbach, the challenged provisions were confined to those re[533]*533gions of the country where voting discrimination had been most flagrant, see 383 U. S., at 315, and affected a discrete class of state laws, i. e., state voting laws. Furthermore, to ensure that the reach of the Voting Rights Act was limited to those cases in which constitutional violations were most likely (in order to reduce the possibility of overbreadth), the coverage under the Act would terminate “at the behest of States and political subdivisions in which the danger of substantial voting discrimination has not materialized during the preceding five years.” Id., at 331. The provisions restricting and banning literacy tests, upheld in Katzenbach v. Morgan, 384 U. S. 641 (1966), and Oregon v. Mitchell, supra, attacked a particular type of voting qualification, one with a long history as a “notorious means to deny and abridge voting rights on racial grounds.” South Carolina v. Katzenbach, 383 U. S., at 355 (Black, J., concurring and dissenting). In City of Rome, supra, the Court rejected a challenge to the constitutionality of a Voting Rights Act provision which required certain jurisdictions to submit changes in electoral practices to the Department of Justice for preimplementation review. The requirement was placed only on jurisdictions with a history of intentional racial discrimination in voting. Id., at 177. Like the provisions at issue in South Carolina v. Katzenbach, this provision permitted a covered jurisdiction to avoid preclearance requirements under certain conditions and, moreover, lapsed in seven years. This is not to say, of course, that § 5 legislation requires termination dates, geographic restrictions, or egregious predicates. Where, however, a congressional enactment pervasively prohibits constitutional state action in an effort to remedy or to prevent unconstitutional state action, limitations of this kind tend to ensure Congress’ means are proportionate to ends legitimate under § 5.
The stringent test RFRA demands of state laws reflects a lack of proportionality or congruence between the means adopted and the legitimate end to be achieved. If an objector can show a substantial burden on his free exercise, the [534]*534State must demonstrate a compelling governmental interest and show that the law is the least restrictive means of furthering its interest. Claims that a law substantially burdens someone’s exercise of religion will often be difficult to contest. See Smith, 494 U. S., at 887 (“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?”); id., at 907 (“The distinction between questions of centrality and questions of sincerity and burden is admittedly fine ...”) (O’Connor, J., concurring in judgment). Requiring a State to demonstrate a compelling interest and show that it has adopted the least restrictive means of achieving that interest is the most demanding test known to constitutional law. If “ ‘compelling interest’ really means what it says . . ., many laws will not meet the test. . . . [The test] would open the prospect of constitutionally required religious. exemptions from civic obligations of almost every conceivable kind.” Id., at 888. Laws valid under Smith would fall under RFRA without regard to whether they had the object of stifling or punishing free exercise. We make these observations not to reargue the position of the majority in Smith but to illustrate the substantive alteration of its holding attempted by RFRA. Even assuming RFRA would be interpreted in effect to mandate some lesser test, say, one equivalent to intermediate scrutiny, the statute nevertheless would require searching judicial scrutiny of state law with the attendant likelihood of invalidation. This is a considerable congressional intrusion into the States’ traditional prerogatives and general authority to regulate for the health and welfare of their citizens.
The substantial costs RFRA exacts, both in practical terms of imposing a heavy litigation burden on the States and in terms of curtailing their traditional general regulatory power, far exceed any pattern or practice of unconstitutional conduct under the Free Exercise Clause as interpreted in Smith. Simply put, RFRA is not designed to identify and counteract state laws likely to be unconstitutional because of [535]*535their treatment of religion. In most cases, the state laws to which RFRA applies are not ones which will have been motivated by religious bigotry. If a state law disproportionately burdened a particular class of religious observers, this circumstance might be evidence of an impermissible legislative motive. Cf. Washington v. Davis, 426 U. S. 229, 241 (1976). RFRA’s substantial-burden test, however, is not even a discriminatory-effects or disparate-impact test. It is a reality of the modern regulatory state that numerous state laws, such as the zoning regulations at issue here, impose a substantial burden on a large class of individuals. When the exercise of religion has been burdened in an incidental way by a law of general application, it does not follow that the persons affected have been burdened any more than other citizens, let alone burdened because of their religious beliefs. In addition, the Act imposes in every case a least restrictive means requirement — a requirement that was not used in the pre-Smit/i jurisprudence RFRA purported to codify — which also indicates that the legislation is broader than is appropriate if the goal is to prevent and remedy constitutional violations.
When Congress acts within its sphere of power and responsibilities, it has not just the right but the duty to make its own informed judgment on the meaning and force of the Constitution. This has been clear from the early days of the Republic. In 1789, when a Member of the House of Representatives objected to a debate on the constitutionality of legislation based on the theory that “it would be officious” to consider the constitutionality of a measure that did not affect the House, James Madison explained that “it is incontrovertibly of as much importance to this branch of the Government as to any other, that the constitution should be preserved entire. It is our duty.” 1 Annals of Congress 500 (1789). Were it otherwise, we would not afford Congress the presumption of validity its enactments now enjoy.
Our national experience teaches that the Constitution is preserved best when each part of the Government respects [536]*536both the Constitution and the proper actions and determinations of the other branches. When the Court has interpreted the Constitution, it has acted within the province of the Judicial Branch, which embraces the duty to say what the law is. Marbury v. Madison, 1 Cranch, at 177. When the political branches of the Government act against the background of a judicial interpretation of the Constitution already issued, it must be understood that in later cases and controversies the Court will treat its precedents with the respect due them under settled principles, including stare decisis, and contrary expectations must be disappointed. RFRA was designed to control cases and controversies, such as the one before us; but as the provisions of the federal statute here invoked are beyond congressional authority, it is this Court’s precedent, not RFRA, which must control.
* * *
It is for Congress in the first instance to “determin[e] whether and what legislation is needed to secure the guarantees of the Fourteenth Amendment,” and its conclusions are entitled to much deference. Katzenbach v. Morgan, 384 U. S., at 651. Congress’ discretion is not unlimited, however, and the courts retain the power, as they have since Marbury v. Madison, to determine if Congress has exceeded its authority under the Constitution. Broad as the power of Congress is under the Enforcement Clause of the Fourteenth Amendment, RFRA contradicts vital principles necessary to maintain separation of powers and the federal balance. The judgment of the Court of Appeals sustaining the Act’s constitutionality is reversed.
It is so ordered.
Justice Scalia joins all but Part III-A-1 of this opinion.