with whom The Chief Justice and Justice Scalia join, dissenting.
The Louisiana statutory scheme the Court strikes down today is not some quirky relic of a bygone age, but a codification of the current provisions of the American Law Institute’s Model Penal Code. Invalidating this quite reasonable scheme is bad enough; even worse is the Court’s failure to explain precisely what is wrong with it. In parts of its opinion, the Court suggests that the scheme is unconstitutional because it provides for the continued confinement of insanity acquittees who, although still dangerous, have “recovered” their sanity. Ante, at 77 (“[T]he committed acquittee is entitled to release when he has recovered his sanity or is no longer dangerous”) (emphasis added; internal quotation marks omitted). In other parts of the opinion, the Court suggests — and the concurrence states explicitly — that the constitutional flaw with this scheme is not that it provides for the confinement of sane insanity acquittees, but that it (allegedly) provides for their “indefinite” confinement in a mental facility. Ante, at 82; ante, at 86-87 (O’Connor, J., concurring in part and concurring in judgment). Nothing in the Constitution, this Court’s precedents, or our society’s traditions authorizes the Court to invalidate the Louisiana scheme on either of these grounds. I would therefore affirm the judgment of the Louisiana Supreme Court.
*103HH
The Court errs, in large part, because it fails to examine in detail the challenged statutory scheme and its application in this ease. Under Louisiana law, a verdict of “not guilty by reason of insanity” differs significantly from a verdict of “not guilty.” A simple verdict of not guilty following a trial means that the State has failed to prove all of the elements of the charged crime beyond a reasonable doubt. See, e. g., State v. Messiah, 538 So. 2d 175, 180 (La. 1988) (citing In re Winship, 397 U. S. 358 (1970)); cf. La. Code Crim. Proc. Ann., Art. 804(A)(1) (West 1969). A verdict of not guilty by reason of insanity, in contrast, means that the defendant committed the crime, but established that he was “incapable of distinguishing between right and wrong” with respect to his criminal conduct. La. Rev. Stat. Ann. §14:14 (West 1986). Insanity, in other words, is an affirmative defense that does not negate the State’s proof, but merely “exempt[s the defendant] from criminal responsibility.” Ibid. As the Louisiana Supreme Court has summarized: “The State’s traditional burden of proof is to establish beyond a reasonable doubt all necessary elements of the offense. Once this rigorous burden of proof has been met, it having been shown that defendant has committed a crime, the defendant . . . bear[s] the burden of establishing his defense of insanity in order to escape punishment.” State v. Marmillion, 339 So. 2d 788, 796 (1976) (emphasis added). See also State v. Surrency, 88 So. 240, 244 (La. 1921).
Louisiana law provides a procedure for a judge to render a verdict of not guilty by reason of insanity upon a plea without a trial. See La. Code Crim. Proc. Ann., Art. 558.1 (West Supp. 1991). The trial court apparently relied on this procedure when it committed Foucha. See 563 So. 2d 1138, 1139, *104n. 3 (La. 1990).1 After ordering two experts to examine Foucha, the trial court issued the following judgment:
“After considering the law and the evidence adduced in this matter, the Court finds that the accused, Terry Foucha, is unable to appreciate the usual, natural and probable consequences of his acts; that he is unable to distinguish right from wrong; that he is a menace to himself and to others; and that he was insane at the time of the commission of the above crimes and that he is presently insane.” App. 6.
After adjudicating a defendant not guilty by reason of insanity, a trial court must hold a hearing on the issue of dangerousness. The law specifies that “[i]f the court determines that the defendant cannot be released without a danger to others or to himself, it shall order him committed to . . . [a] mental institution.” La. Code Crim. Proc. Ann., Art. 654 (West Supp. 1991).2 “ ‘Dangerous to others’ means *105the condition of a person whose behavior or significant threats support a reasonable expectation that there is a substantial risk that he will inflict physical harm upon another person in the near future.” La. Rev. Stat. Ann. §28:2(3) (West 1986) (emphasis added). “ ‘Dangerous to self’ means the condition of a person whose behavior, significant threats or inaction supports a reasonable expectation that there is a substantial risk that he will inflict physical or severe emotional harm upon his own person.” §28:2(4).
After holding the requisite hearings, the trial court in this case ordered Foucha committed to the Feliciana Forensic Facility. After his commitment, Foucha was entitled, upon request, to another hearing six months later and at yearly intervals after that. See La. Code Crim. Proc. Ann., Art. 655(B) (West Supp. 1991).3 In addition, Louisiana law provides that a release hearing must be held upon recommendation by the superintendent of a mental institution. See Art. 655(A).4 In early 1988, Feliciana’s superintendent recom*106mended that Foucha be released, and a three-doctor panel met to review the case. On March 21, 1988, the panel issued a report pursuant to Article 656.5 The panel concluded that “there is no evidence of mental illness.” App. 10. In fact, the panel stated that there was “never any evidence of mental illness or disease since admission.” Ibid, (emphasis added). Although the panel did not discuss whether Foucha was dangerous, it recommended to the trial court that he be conditionally released.
As a result of these recommendations, the trial court scheduled a hearing to determine whether Foucha should be released. Under La. Code Crim. Proc. Ann., Art. 657 (West Supp. 1991),6 Foucha had the burden at this hearing to prove *107that he could be released without danger to others or to himself. The court appointed two experts (the same doctors who had examined Foucha at the time of his original commitment) to evaluate his dangerousness. These doctors concluded that Foucha “is presently in remission from mental illness,” but said that they could not “certify that he would not constitute a menace to himself or to others if released.” App. 12. On November 29, 1988, the trial court held the hearing, at which Foucha was represented by counsel. The court concluded that Foucha “is a danger to himself, and to others,” id., at 24, and ordered that he be returned to Feliciana.7
II
The Court today concludes that Louisiana has denied Foucha both procedural and substantive due process. In my view, each of these conclusions is wrong. I shall discuss them in turn.
A
What the Court styles a “procedural” due process analysis is in reality an equal protection analysis. The Court first asserts (contrary to state law) that Foucha cannot be held as an insanity acquittee once he “becomes” sane. Ante, at 78-79. *108That being the case, he is entitled to the same treatment as civil committees. “[I]f Foucha can no longer be held as an insanity acquittee,” the Court says, “he is entitled to constitutionally adequate procedures [those afforded in civil commitment proceedings] to establish the grounds for his confinement.” Ante, at 79 (emphasis added). This, of course, is an equal protection argument (there being no rational distinction between A and B, the State must treat them the same); the Court does not even pretend to examine the fairness of the release procedures the State has provided.
I cannot agree with the Court’s conclusion because I believe that there is a real and legitimate distinction between insanity acquittees and civil committees that justifies procedural disparities. Unlike civil committees, who have not been found to have harmed society, insanity acquittees have been found in a judicial proceeding to have committed a criminal act.
That distinction provided the ratio decidendi for our most relevant precedent, Jones v. United States, 463 U. S. 354 (1983). That case involved a man who had been automatically committed to a mental institution after being acquitted of a crime by reason of insanity in the District of Columbia (i. e., he had not been given the procedures afforded to civil committees). We rejected both of his procedural due process challenges to his commitment. First, we held that an insanity acquittal justified automatic commitment of the ac-quittee (even though he might presently be sane), because Congress was entitled to decide that the verdict provided a reasonable basis for inferring dangerousness and insanity at the time of commitment. Id., at 366. The Government’s interest in avoiding a de novo commitment hearing following every insanity acquittal, we said, outweighed the acquittee’s interest in avoiding unjustified institutionalization. Ibid. Second, we held that the Constitution did not require, as a predicate for the indefinite commitment of insanity acquit-tees, proof of insanity by “clear and convincing” evidence, as *109required for civil committees by Addington v. Texas, 441 U. S. 418 (1979). There are, we recognized, “important differences between the class of potential civil-commitment candidates and the class of insanity acquittees that justify differing standards of proof.” Jones, 463 U. S., at 367. In sharp contrast to a civil committee, an insanity acquittee is institutionalized only where “the acquittee himself advances insanity as a defense and proves that his criminal act was a product of his mental illness,” and thus “there is good reason for diminished concern as to the risk of error.” Ibid, (emphasis in original). “More important, the proof that he committed a criminal act . . . eliminates the risk that he is being committed for mere ‘idiosyncratic behavior.’” Ibid. Thus, we concluded, the preponderance of the evidence standard comports with due process for commitment of insanity acquit-tees. Id., at 368. “[I]nsanity acquittees constitute a special class that should be treated differently from other candidates for commitment.” Id., at 370.
The Court today attempts to circumvent Jones by declaring that a State’s interest in treating insanity acquittees differently from civil committees evaporates the instant an acquittee “becomes sane.” I do not agree. As an initial matter, I believe that it is unwise, given our present understanding of the human mind, to suggest that a determination that a person has “regained sanity” is precise. “Psychiatry is not... an exact science, and psychiatrists disagree widely and frequently on what constitutes mental illness.” Ake v. Oklahoma, 470 U. S. 68, 81 (1985). Indeed,
“[w]e have recognized repeatedly the ‘uncertainty of diagnosis in this field and the tentativeness of professional judgment. The only certain thing that can be said about the present state of knowledge and therapy regarding mental disease is that science has not reached finality of judgment.’ The lesson we have drawn is not that government may not act in the face of this uncertainty, but rather that courts should pay particular *110deference to reasonable legislative judgments.” Jones, supra, at 365, n. 13 (quoting Greenwood v. United States, 350 U. S. 366, 375 (1956); citations omitted).
In this very case, the panel that evaluated Foucha in 1988 concluded that there was “never any evidence of mental illness or disease since admission,” App. 10; the trial court, of course, concluded that Foucha was “presently insane,” id., at 6, at the time it accepted his plea and sent him to Feliciana.
The distinction between civil committees and insanity ac-quittees, after all, turns not on considerations of present sanity, but instead on the fact that the latter have “already unhappily manifested the reality of anti-social conduct,” Dixon v. Jacobs, 138 U. S. App. D. C. 319, 33'4, 427 F. 2d 589, 604 (1970) (Leventhal, J., concurring). “[TJhe prior anti-social conduct of an insanity acquittee justifies treating such a person differently from ones otherwise civilly committed for purposes of deciding whether the patient should be released.” Powell v. Florida, 579 F. 2d 324, 333 (CA5 1978) (emphasis added); see also United States v. Ecker, 177 U. S. App. D. C. 31, 50, 543 F. 2d 178, 197 (1976), cert, denied, 429 U. S. 1063 (1977). While a State may renounce a punitive interest by offering an insanity defense, it does not follow that, once the acquittee’s sanity is “restored,” the State is required to ignore his criminal act, and to renounce all interest in protecting society from him. “The state has a substantial interest in avoiding premature release of insanity acquittees, who have committed acts constituting felonies and have been declared dangerous to society.” Hickey v. Morris, 722 F. 2d 543, 548 (CA9 1983).
Furthermore, the Federal Constitution does not require a State to “ignore the danger of ‘calculated abuse of the insanity defense.’” Warren v. Harvey, 632 F. 2d 925, 932 (CA2 1980) (quoting United States v. Brown, 155 U. S. App. D. C. 402, 407, 478 F. 2d 606, 611 (1973)). A State that decides to offer its criminal defendants an insanity defense, which the defendant himself is given the choice of invoking, is surely *111allowed to attach to that defense certain consequences that prevent abuse. Cf. Lynch v. Overholser, 369 U. S. 705, 715 (1962) (“Congress might have considered it appropriate to provide compulsory commitment for those who successfully invoke an insanity defense in order to discourage false pleas of insanity”).
“In effect, the defendant, by raising the defense of insanity — and he alone can raise it — postpones a determination of his present mental health and acknowledges the right of the state,, upon accepting his plea, to detain him for diagnosis, care, and custody in a mental institution until certain specified conditions are met. . . . [Commitment via the criminal process . . . thus is more akin to ‘voluntary’ than ‘involuntary’ civil commitment.” Goldstein & Katz, Dangerousness and Mental Illness, Some Observations on the Decision to Release Persons Acquitted by Reason of Insanity, 70 Yale L. J. 225, 230 (1960) (footnote omitted).
A State may reasonably decide that the integrity of an insanity-acquittal scheme requires the continued commitment of insanity acquittees who remain dangerous. Surely, the citizenry would not long tolerate the insanity defense if a serial killer who convinces a jury that he is not guilty by reason of insanity is returned to the streets immediately after trial by convincing a different factfinder that he is not in fact insane.
As the American Law Institute has explained:
“It seemed preferable to the Institute to make dangerousness the criterion for continued custody, rather than to provide that the committed person may be discharged or released when restored to sanity as defined by the mental hygiene laws. Although his mental disease may have greatly improved, [an insanity acquittee] may still be dangerous because of factors in his personality and ■background other than mental disease. Also, such a *112standard provides a means for the control of the occasional defendant who may be quite dangerous but who successfully feigned mental disease to gain an acquittal.” Model Penal Code §4.08, Comment 3, pp. 259-260 (1985).8
That this is a reasonable legislative judgment is underscored by the fact that it has been made by no fewer than 11 state legislatures, in addition to Louisiana’s, which expressly provide that insanity acquittees shall not be released as long as they are dangerous, regardless of sanity.9
*113The Court suggests an alternative “procedural” due process theory that is, if anything, even less persuasive than its principal theory. “[K]eeping Foucha against his will in a mental institution is improper absent a determination in civil commitment proceedings of current mental illness and dangerousness.” Ante, at 78 (emphasis added). The Court cites Vitek v. Jones, 445 U. S. 480 (1980), as support. There are two problems with this theory. First, it is illogical: Louisiana cannot possibly extend Foucha’s incarceration by adding the procedures afforded to civil committees, since it is impossible to civilly commit someone who is not presently *114mentally ill. Second, the theory is not supported by Vitek. Stigmatization (our concern in Vitek) is simply not a relevant consideration where insanity acquittees are involved. As we explained in Jones: “A criminal defendant who successfully raises the insanity defense necessarily is stigmatized by the verdict itself, and thus the commitment causes little additional harm in this respect.” 463 U. S., at 367, n. 16; see also Warren v. Harvey, 632 F. 2d, at 931-932. (This is in sharp contrast to situations involving civil committees. See Addington, 441 U. S., at 425-426; Vitek, supra, at 492-494.) It is implausible, in my view, that a person who chooses to plead not guilty by reason of insanity and then spends several years in a mental institution becomes unconstitutionally stigmatized by continued confinement in the institution after “regaining” sanity.
In my view, there was no procedural due process violation in this case. Articles 654, 655, and 657 of the Louisiana Code of Criminal Procedure, as noted above, afford insanity acquittees the opportunity to obtain release by demonstrating at regular intervals that they no longer pose a threat to society. These provisions also afford judicial review of such determinations. Pursuant to these procedures, and based upon testimony of experts, the Louisiana courts determined not to release Foucha at this time because the evidence did not show that he ceased to be dangerous. Throughout these proceedings, Foucha was represented by state-appointed counsel. I see no plausible argument that these procedures denied Foucha a fair hearing on the issue involved or that Foucha needed additional procedural protections.10 See Mathews v. Eldridge, 424 U. S. 319 (1976); Patterson v. New York, 432 U. S. 197 (1977); cf. Addington, supra, at 427-432; *115Jones, supra, at 363-368; Benham v. Ledbetter, 785 F. 2d 1480, 1486-1488 (CA11 1986).11
B
The Court next concludes that Louisiana’s statutory scheme must fall because it violates Foucha’s substantive due process rights. Ante, at. 80-83, and n. 6. I disagree. Until today, I had thought that the analytical framework for evaluating substantive due process claims was relatively straightforward. Certain substantive rights we have recognized as “fundamental”; legislation trenching upon these is subjected to “strict scrutiny,” and generally will be invalidated unless the State demonstrates a compelling interest and narrow tailoring. Such searching judicial review of state legislation, however, is the exception, not the rule, in our democratic and federal system; we have consistently emphasized that “the Court has no license to invalidate legislation which it thinks merely arbitrary or unreasonable.” Regents of University of Michigan v. Ewing, 474 U. S. 214, 226 (1985) (internal quotation marks omitted). Except in the unusual case where a fundamental right is infringed, then, federal judicial scrutiny of the substance of state legislation under the Due Process Clause of the Fourteenth Amendment is not exacting. See, e. g., Bowers v. Hardwick, 478 U. S. 186, 191-196 (1986).
In striking down Louisiana’s scheme as a violation of substantive rights guaranteed by the Due Process Clause, the *116Court today ignores this well-established analytical framework. First, the Court never explains whether we are dealing here with a fundamental right, and, if so, what right. Second, the Court never discloses what standard of review applies. Indeed, the Court’s opinion is contradictory on both these critical points.
As to the first point: The Court begins its substantive due process analysis by invoking the substantive right to “[freedom from bodily restraint.” Ante, at 80. Its discussion then proceeds as if the problem here is that Foucha, an insanity acquittee, continues to be confined after recovering his sanity, ante, at 80-81; thus, the Court contrasts this case to United States v. Salerno, 481 U. S. 739 (1987), a case involving the confinement of pretrial detainees. But then, abruptly, the Court shifts liberty interests. The liberty interest at stake here, we are told, is not a liberty interest in being free “from bodily restraint,” but instead the more specific (and heretofore unknown) “liberty interest under the Constitution in being freed from, [1] indefinite confinement [2] in a mental facility” Ante, at 82 (emphasis added). See also ante, at 86-87 (O’Connor, J., concurring in part and concurring in judgment). So the problem in this case is apparently not that Louisiana continues to confine insanity acquittees who have “become” sane (although earlier in the opinion the Court interprets our decision in Jones as having held that such confinement is unconstitutional, see ante, at 77-78), but that under Louisiana law, “sane” insanity acquit-tees may be held “indefinitely” “in a mental facility.”
As to the second point: “A dispute regarding the appropriate standard of review may strike some as a lawyers’ quibble over words, but it is not.” Metro Broadcasting, Inc. v. FCC, 497 U. S. 547, 610 (1990) (O’Connor, J., dissenting). The standard of review determines when the-Due Process Clause of the Fourteenth Amendment will override a State’s substantive policy choices, as reflected in its laws. The Court initially says that “[d]ue process requires that the na*117ture of commitment bear some reasonable relation to the purpose for which the individual is committed.” Ante, at 79 (emphasis added). Later in its opinion, however, the Court states that.the Louisiana scheme violates substantive due process not because it is not “reasonably related” to the State’s purposes, but instead because its detention provisions are not “sharply focused” or “carefully limited,” in contrast to the scheme we upheld in Salerno. Ante, at 81. Does that mean that the same standard of review applies here that we applied in Salerno, and, if so, what is that standard? The Court quite pointedly avoids answering these questions. Similarly, Justice O’Connor does not reveal exactly what standard of review she believes applicable, but appears to advocate a heightened standard heretofore unknown in our case law. Ante, at 87-88 (“It might therefore be permissible for Louisiana to confine an insanity acquittee who has regained sanity if... the nature and duration of detention were tailored to reflect pressing public safety concerns related to the acquittee’s continuing dangerousness” (emphasis added)).
To the extent the Court invalidates the Louisiana scheme on the ground that it violates some general substantive due process right to “freedom from bodily restraint” that triggers strict scrutiny, it is wrong — and dangerously so. To the extent the Court suggests that Louisiana has violated some more limited right to freedom from indefinite commitment in a mental facility (a right, by the way, never asserted by Foucha in this or any other court) that triggers some unknown standard of review, it is also wrong. I shall discuss these two possibilities in turn.
1
I fully agree with the Court, ante, at 80, and with Justice Kennedy, ante, at 90, that freedom from involuntary confinement is at the heart of the “liberty” protected by the Due Process Clause. But a liberty interest per se is not the same thing as a fundamental right. Whatever the exact scope of *118the fundamental right to “freedom from bodily restraint” recognized by our cases,12 it certainly cannot be defined at the exceedingly great level of generality the Court suggests today. There is simply no basis in our society’s history or in the precedents of this Court to support the existence of a sweeping, general fundamental right to “freedom from bodily restraint” applicable to all persons in all contexts. If convicted prisoners could claim such a right, for example, we would subject all prison sentences to strict scrutiny. This we have consistently refused to do. See, e. g., Chapman v. United States, 500 U. S. 453, 465 (1991).13
The critical question here, then, is whether insanity ac-quittees have a fundamental right to “freedom from bodily *119restraint” that triggers strict scrutiny of their confinement. Neither Foucha nor the Court provides any evidence that our society has ever recognized any such right. To the contrary, historical evidence shows that many States have long provided for the continued institutionalization of insanity acquittees who remain dangerous. See, e. g., H. Weihofen, Insanity as a Defense in Criminal Law 294-332 (1933); A. Goldstein, The Insanity Defense 148-149 (1967).
Moreover, this Court has never applied strict scrutiny to the substance of state laws involving involuntary confinement of the mentally ill, much less to laws involving the confinement of insanity acquittees. To the contrary, until today we have subjected the substance of such laws only to very deferential review. Thus, in Jackson v. Indiana, 406 U. S. 715, 738 (1972), we held that Indiana’s provisions for the indefinite institutionalization of incompetent defendants violated substantive due process because they did not bear any “reasonable” relation to the purpose for which the defendant was committed. Similarly, in O’Connor v. Donaldson, 422 U. S. 563 (1975), we held that the confinement of a nondan-gerous mentally ill person was unconstitutional not because the State failed to show a compelling interest and narrow tailoring, but because the State had no legitimate interest whatsoever to justify such confinement. See id., at 575-576. See also id., at 580 (Burger, C. J., concurring) (“Commitment must be justified on the basis of a legitimate state interest, and the reasons for committing a particular individual must be established in an appropriate proceeding. Equally important, confinement must cease when those reasons no longer exist” (emphasis added)).
Similarly, in Jones, we held (in addition to the procedural due process holdings described above) that there was no substantive due process bar to holding an insanity acquittee beyond the period for which he could have been incarcerated if convicted. We began by explaining the standard for our analysis: “The Due Process Clause ‘requires that the nature *120and duration of commitment bear some reasonable relation to the purpose for which the individual is committed.’” 463 U. S., at 368 (emphasis added) (quoting Jackson, supra, at 738). We then held that “[i]n light of the congressional purposes underlying commitment of insanity acquittees [in the District of Columbia,]” which we identified as treatment of the insanity acquittee’s mental illness and protection of the acquittee and society, “petitioner clearly errs in contending that an acquittee’s hypothetical maximum sentence provides the constitutional limit for his commitment.” 463 U. S., at 368 (emphasis added). Given that the commitment law was reasonably related to Congress’ purposes, this Court had no basis for invalidating it as a matter of substantive due process.
It is simply wrong for the Court to assert today that we “held” in Jones that “‘the committed acquittee is entitled to release when he has recovered his sanity or is no longer dangerous.’ ” Ante, at 77 (quoting Jones, 463 U. S., at 368).14 We specifically noted in Jones that no issue regarding the standards for the release of insanity acquittees was before us. Id., at 363, n. 11. The question we were answering in the part of Jones from which the Court quotes was whether it is permissible to hold an insanity acquittee for a period longer than he could have been incarcerated if convicted, not whether it is permissible to hold him once he becomes “sane.” As noted above, our substantive due process analysis in Jones was straightforward: Did the means chosen by Congress (commitment of insanity acquittees until *121they have recovered their sanity or are no longer dangerous) reasonably fit Congress’ ends (treatment of the acquit-tee’s mental illness and protection, of society from his dangerousness)?15
In its arguments before this Court, Louisiana chose to place primary reliance on our decision in United States v. Salerno, 481 U. S. 739 (1987), in which we upheld provisions of the Bail Reform Act of 1984 that allowed limited pretrial detention of criminal suspects. That case, as the Court notes, ante, at 81-83, is readily distinguishable. Insanity ac-quittees, in sharp and obvious contrast to pretrial detainees, have had their day in court. Although they have not been convicted of crimes, neither have they been exonerated, as they would have been upon a determination of “not guilty” simpliciter. Insanity acquittees thus stand in a fundamentally different position from persons who have not been adjudicated to have committed criminal acts. That is what distinguishes this case (and what distinguished Jones) from Salerno and Jackson v. Indiana, supra. In Jackson, as in Salerno, the State had not proved beyond a reasonable doubt that the accused had committed criminal acts or otherwise was dangerous. See Jones, supra, at 364, n. 12. The Court disregards this critical distinction, and apparently deems applicable the same scrutiny to pretrial de*122tainees as to persons determined in a judicial proceeding to have committed a criminal act.16
If the Court indeed means to suggest that all restrictions on “freedom from bodily restraint” are subject to strict scrutiny, it has (at a minimum) wrought a revolution in the treatment of the mentally ill. Civil commitment as we know it would almost certainly be unconstitutional; only in the rarest of circumstances will a State be able to show a “compelling interest,” and one that can be served in no other way, in involuntarily institutionalizing a person. All procedures involving the confinement of insanity acquittees and civil committees would require revamping to meet strict scrutiny. Thus, to take one obvious example, the automatic commitment of insanity acquittees that we expressly upheld in Jones would be clearly unconstitutional, since it is inconceivable that such commitment of persons who may well presently be sane and nondangerous could survive strict scrutiny. (In Jones, of course, we applied no such scrutiny; we upheld the practice not because it was justified by a compelling in*123terest, but because it was based on reasonable legislative inferences about continuing insanity and dangerousness.)
2
As explained above, the Court’s opinion is profoundly ambiguous on the central question in this case: Must the State of Louisiana release Terry Foucha now that he has “regained” his sanity? In other words, is the defect in Louisiana’s statutory scheme that it provides for the confinement of insanity acquittees who have recovered their sanity, or instead that it allows the State to confine sane insanity acquittees (1) indefinitely (2) in a mental facility? To the extent the Court suggests the former, I have already explained why it is wrong. I turn now to the latter possibility, which also is mistaken.
To begin with, I think it is somewhat misleading to describe Louisiana’s scheme as providing for the “indefinite” commitment of insanity acquittees. As explained above, insanity acquittees are entitled to a release hearing every year at their request, and at any . time at the request of a facility superintendent. Like the District of Columbia statute at issue in Jones, then, Louisiana’s statute provides for “indefinite” commitment only to the extent that an acquittee is unable to satisfy the substantive standards for release. If the Constitution did not require a cap on the acquittee’s confinement in Jones, why does it require one here? The Court and Justice . O’Connor have no basis for suggesting that either this Court or the society of which it is a part has recognized some general fundamental right to “freedom from indefinite commitment.” If that were the case, of course, Jones would have involved strict scrutiny and is wrongly decided.
Furthermore, any concerns about “indefinite” commitment here are entirely hypothetical and speculative. Foucha has been confined for eight years. Had he been convicted of the crimes with which he was charged, he could have been incar*124cerated for 32 years. See La. Rev. Stat. Ann. §§ 14:60, 14:94 (West 1986). Thus, I find quite odd Justice O’Connor’s suggestion, ante, at 89, that this case might be different had Louisiana, like the State of Washington, limited confinement to the period for which a defendant might have been imprisoned if convicted. Foucha, of course, would be in precisely the same position today — and for the next 24 years — had the Louisiana statute included such a cap. Thus, the Court apparently finds fault with the Louisiana statute not because it has been applied to Foucha in an unconstitutional manner, but because the Court can imagine it being applied to someone else in an unconstitutional manner. That goes against the first principles of our jurisprudence. See, e. g., Salerno, 481 U. S., at 745 (“The fact that [a detention statute] might operate unconstitutionally under some conceivable set of circumstances is insufficient to render it wholly invalid, since we have not recognized an ‘overbreadth’ doctrine outside the limited context of the First Amendment”).17
Finally, I see no basis for holding that the Due Process Clause per se prohibits a State from continuing to confine in a “mental institution” — the federal constitutional definition of which remains unclear — an insanity acquittee who has recovered his sanity. As noted above, many States have long provided for the continued detention of insanity acquittees who remain dangerous. Neither Foucha nor the Court present any evidence that these States have traditionally transferred such persons from mental institutions to other detention facilities. Therefore, there is simply no basis for this Court to recognize a “fundamental right” for a sane insanity acquittee to be transferred out of a mental facility. “In an attempt to limit and guide interpretation of the [Due Process] Clause, we have insisted not merely that the interest *125denominated as a ‘liberty’ be ‘fundamental’ (a concept that, in isolation, is hard to objectify), but also that it be an interest traditionally protected by our society.” Michael H. v. Gerald D., 491 U. S. 110, 122 (1989) (plurality opinion).
Removing sane insanity acquittees from mental institutions may make eminent sense as a policy matter, but the Due Process Clause does not require the States to conform to the policy preferences of federal judges. “The Court is most vulnerable and comes nearest to illegitimacy when it deals with judge-made constitutional law having little or no cognizable roots in the language or design of the Constitution.” Bowers, 478 U. S., at 194. I have no idea what facilities the Court or Justice O’Connor believe the Due Process Clause mandates for the confinement of sane-but-dangerous insanity acquittees. Presumably prisons will not do, since imprisonment is generally regarded as “punishment.” May a State designate a wing of a mental institution or prison for sane insanity acquittees? May a State mix them with other detainees? Neither the Constitution nor our society’s traditions provide any answer to these questions.18
3
“So-called ‘substantive due process’ prevents the government from engaging in conduct that ‘shocks the conscience,’ Rochin v. California, 342 U. S. 165, 172 (1952), or interferes *126with rights ‘implicit in the concept of ordered liberty.' Palko v. Connecticut, 302 U. S. 319, 325-326 (1937).” Salerno, supra, at 746. The legislative scheme the Court invalidates today is, at the very least, substantively reasonable. With all due respect, I do not remotely think it can be said that the laws in question “offen[d] some principle of justice so rooted in the traditions and conscience of our people as to be ranked as fundamental.” Snyder v. Massachusetts, 291 U. S. 97, 105 (1934). Therefore, in my view, this Court is not entitled, as a matter of substantive due process, to strike them down.
I respectfully dissent.
Under La. Code Crim. Proc. Ann., Art. 558.1 (West Supp. 1991), a criminal defendant apparently concedes that he committed the crime, and advances his insanity as the sole ground on which to avoid conviction. Fou-cha does not challenge the procedures whereby he was adjudicated not guilty by reason of insanity; nor does he deny that he committed the crimes with which he was charged.
Article 664 provides in pertinent part:
“When a defendant is found not guilty by reason of insanity in any [non-capital] felony case, the court shall remand him to the parish jail or to a private mental institution approved by the court and shall promptly hold a contradictory hearing at which the defendant shall have the burden of proof, to determine whether the defendant can be discharged or can be released on probation, without danger to others or to himself. If the court determines that the defendant cannot be released without danger to others or to himself, it shall order him committed to a proper state mental institution or to a private mental institution approved by the court for custody, care, and treatment. If the court determines that the defendant can be discharged or released on probation without danger to others or to himself, the court shall either order his discharge, or order his release on probation subject to specified conditions for a fixed or an indeterminate period. The court shall assign written findings of fact and conclusions of *105law; however, the assignment of reasons shall not delay the implementation of judgment.”
Article 655(B) provides:
“A person committed pursuant to Article 654 may make application to the review panel for discharge or for release on probation. Such application by a committed person may not be filed until the committed person has been confined for a period of at least six months after the original commitment. If the review panel recommends to the court that the person be discharged, conditionally or unconditionally, or placed on probation, the court shall conduct a hearing following notice to the district attorney. If the recommendation of the review panel or the court is adverse, the applicant shall not be permitted to file another application until one year has elapsed from the date of determination.”
Article 655(A) provides:
“When the superintendent of a mental institution is of the opinion that a person committed pursuant to Article 654 can be discharged or can be released on probation, without danger to others or to himself, he shall recommend the discharge or release of the person in a report to a review panel comprised of the person’s treating physician, the clinical director of the facility to which the person is committed, and a physician or psychologist who served on the sanity commission which recommended commitment of the person. If any member of the panel is unable to serve, a *106physician or a psychologist engaged in the practice of clinical or counseling psychology with at least three years’ experience in the field of mental health shall be appointed by the remaining members. The panel shall review all reports received promptly. After review, the panel shall make a recommendation to the court by which the person was committed as to the person’s mental condition and whether he can be discharged, conditionally or unconditionally, or placed on probation, without being a danger to others or himself. If the review panel recommends to the court that the person be discharged, conditionally or unconditionally, or placed on probation, the court shall conduct a contradictory hearing following notice to the district attorney.”
Article 656 provides:
“A. Upon receipt of the superintendent’s report, filed in conformity with Article 665, the review panel may examine the committed person and report, to the court promptly, whether he can be safely discharged, conditionally or unconditionally, or be safely released on probation, without danger to others or to himself
“B. The committed person or the district attorney may also retain a physician to examine the committed person for the same purpose. The physician’s report shall be filed with the court.”
Article 657 provides:
“After considering the report or reports filed pursuant to Articles 655 and 656, the court may either continue the commitment or hold a contradictory hearing to determine whether the committed person can be discharged, or can be released on probation, without danger to others or to *107himself. At the hearing the burden shall be upon the committed person to prove that he can be discharged, or can be released on probation, without danger to others or to himself. After the hearing, and upon filing written findings of fact and conclusions of law, the court may order the committed person discharged, released on probation subject to specified conditions for a fixed or an indeterminate period, or recommitted to the state mental institution. Notice to the counsel for the committed person and the district attorney of the contradictory hearing shall be given at least thirty days prior to the hearing.”
The Louisiana Supreme Court concluded that the trial court did not abuse its discretion in finding that Foucha had failed to prove that he could be released without danger to others or to himself under La. Code Crim. Proc. Ann., Art. 657 (West Supp. 1991). See 563 So. 2d 1138, 1141 (1990). That issue is not now before us.
The relevant provision of the Model Penal Code, strikingly similar to Article 667 of the Louisiana Code of Criminal Procedure, see supra, n. 6, provides in part as follows:
“If the Court is satisfied by the report filed pursuant to Subsection (2) of this Section and such testimony of the reporting psychiatrists as the Court deems necessary that the committed person may be discharged or released on condition without danger to himself or others, the Court shall order his discharge or his release on such conditions as the Court determines to be necessary. If the Court is not so satisfied, it shall promptly order a hearing to determine whether such person may safely be discharged or released. Any such hearing shall be deemed a civil proceeding and the burden shall be upon the committed person to prove that he may safely be discharged or released.” Model Penal Code §4.08(3) (Proposed Official Draft 1962).
See Cal. Penal Code Ann. § 1026.2(e) (West Supp. 1992) (insanity aequit-tee not entitled to release until court determines that he “will not be a danger to the health and safety of others, including himself”); Del. Code Ann., Tit. 11, § 403(b) (1987) (insanity aequittee shall be kept institutionalized until court “is satisfied that the public safety will not be endangered by his release”); Haw. Rev. Stat. § 704-415 (1985) (insanity aequittee not entitled to release until court satisfied that aequittee “may safely be discharged or released”); Iowa Rule Crim. Proc. 21.8(e) (insanity aequittee not entitled to release as long as “court finds that continued custody and treatment are necessary to protect the safety of the [acquittee’s] self or others”); Kan. Stat. Ann. § 22-3428(3) (Supp. 1990) (insanity aequittee not entitled to release until “the court finds by clear and convincing evidence that [he] will not be likely to cause harm to self or others if released or discharged”); Mont. Code Ann. §46-14-301(3) (1991) (insanity aequittee not entitled to release until he proves that he “may safely be released”); *113N. J. Stat. Ann. §2C:4-9 (West 1982) (insanity acquittee not entitled to release or discharge until court satisfied that he is not “danger to himself or others”); N. C. Gen. Stat. § 122C-268.1(i) (Supp. 1991) (insanity acquit-tee not entitled to release until he “prove[s] by a preponderance of the evidence that he is no longer dangerous to others”); Va. Code Ann. § 19.2-181(3) (1990) (insanity acquittee not entitled to release until he proves “that he is not'insane or mentally retarded and that his discharge would not be dangerous to the public peace and safety or to himself” (emphasis added)); Wash. Rev. Code § 10.77.200(2) (1990) (“The burden of proof [at a release hearing] shall be upon the [insanity acquittee] to show by a preponderance of the evidence that [he] may be finally discharged without substantial danger to other persons, and without presenting a substantial likelihood of committing felonious acts jeopardizing public safety or security”); Wis. Stat. §971.17(4) (Supp. 1991) (insanity acquittee not entitled to release where court “finds by clear and convincing evidence that the [acquit-tee] would pose a significant risk of bodily harm to himself or herself or to others of serious property damage if conditionally released”).
The Court and the concurrence dispute this list of statutes. Ante, at 84-85, n. 6; ante, at 89 (O’Connor, J., concurring in part and concurring in judgment). They note that two of the States have enacted new laws, not yet effective, modifying their current absolute prohibitions on the release of dangerous insanity acquittees; that courts in two other States have apparently held that mental illness is a prerequisite to confinement; and that three of the States place caps of some sort on the duration of the confinement offinsanity acquittees. Those criticisms miss my point. I cite the 11 state statutes above only to show that the legislative judgments underlying Louisiana’s scheme are far from unique or freakish, and that there is no well-established practice in our society, .either past or present, of automatically releasing sane-but-dangerous insanity acquittees.
Foucha has not argued that the State’s procedures, as applied, are a sham. This would be a different case if Foucha had established that the statutory mechanisms for release were nothing more than window dressing, and that the State in fact confined insanity acquittees indefinitely without meaningful opportunity for review and release.
As explained above, the Court’s “procedural” due process analysis is essentially an equal protection analysis: The Court first disregards the differences between “sane” insanity acquittees and civil committees, and then simply asserts that Louisiana cannot deny Foucha the procedures it gives civil committees. A plurality repeats this analysis in its cumulative equal protection section. See ante, at 84-86. As explained above, I believe that there are legitimate differences between civil committees and insanity acquittees, even after the latter have “become” sane. Therefore, in my view, Louisiana has not denied Foucha equal protection of the laws. Cf. Jones v. United States, 463 U. S. 354, 362, n. 10 (1983).
The Court cites only Youngberg v. Romeo, 457 U. S. 307, 316 (1982), in support of its assertion that “[fjreedom from bodily restraint has always been at the core of the liberty protected by the Due Process Clause from arbitrary governmental action,” ante, at 80. What “freedom from bodily restraint” meant in that case, however, is completely different from what the Court uses the phrase to mean here. Youngberg involved the substantive due process rights of an institutionalized, mentally retarded patient who had been restrained by shackles placed on his arms for portions of each day. See 457 U. S., at 310, and n. 4. What the Court meant by “freedom from bodily restraint,” then, was quite literally freedom not to be physically strapped to a bed. That case in no way established the broad “freedom from bodily restraint” — apparently meaning freedom from all involuntary confinement — that the Court discusses today.
Unless the Court wishes to overturn this line of cases, its substantive due process analysis must rest entirely on the fact that an insanity acquit-tee has not been convicted of a crime. Conviction is, of course, a significant event. But I am not sure that it deserves talismanic significance. Once a State proves beyond a reasonable doubt that an individual has committed a crime, it is, at a minimum, not obviously a matter of federal constitutional concern whether the State proceeds to label that individual “guilty,” “guilty but insane,” or “not guilty by reason of insanity.” A State may just as well decide to label its verdicts “A,” “B,” and “C.” It is surely rather odd to have rules of federal constitutional law turn entirely upon the label chosen by a State. Cf. Railway Express Agency, Inc. v. Virginia, 358 U. S. 434, 441 (1969) (constitutionality of state action should not turn on “magic words”).
If this were really a “holding” of Jones, then I am at a loss to understand Justice O’ConnOR’s assertion that the Court today does not hold “that Louisiana may never confine dangerous insanity acquittees after they regain mental health.” Ante, at 87. Either it is true that, as a matter of substantive due process, an insanity acquittee is “‘entitled to release when he has recovered his sanity,’” ante, at 77 (quoting Jones, 463 U. S., at 368), or it is not. The Court apparently cannot make up its mind.
As may be apparent from the discussion in text, we have not been entirely precise as to the appropriate standard of review of legislation in this area. Some of our cases (e. g., O’Connor v. Donaldson, 422 U. S. 563 (1975)) have used the language of rationality review; others (e. g., Jackson v. Indiana, 406 U. S. 715 (1972)) have used the language of “reasonableness,” which may imply a somewhat heightened standard; still others (e. g., Jones) have used the language of both rationality and reasonableness. What is clear from our cases is that the appropriate scrutiny is highly deferential, not strict. We need not decide in this case which precise standard is applicable, since the laws under attack here are at the very least reasonable.
The Court asserts that the principles set forth in this dissent necessarily apply not only to insanity acquittees, but also to convicted prisoners. “Justice Thomas’ rationale for continuing to hold the insanity acquittee would surely justify treating the convicted felon in the same way, and if put to it, it appears that he would permit it.” Ante, at 83, n. 6. That is obviously not so. If Foucha had been convicted of the crimes with which he was charged and sentenced to the statutory maximum of 32 years in prison, the State would not be entitled to extend his sentence at the end of that period. To do so would obviously violate the prohibition on ex post facto laws set forth in Art. I, § 10, cl. 1. But Foucha was not sentenced to incarceration for any definite period of time; to the contrary, he pleaded not guilty by reason of insanity and was ordered institutionalized until he was able to meet the conditions statutorily prescribed for his release. To acknowledge, as I do, that it is constitutionally permissible for a State to provide for the continued confinement of an insanity acquittee who remains dangerous is obviously quite different than to assert that the State is allowed to confine anyone who is dangerous for as long as it wishes.
I fully agree with Justice O’Connor, ante, at 88, that there would be a serious question of rationality had Louisiana sought to institutionalize a sane insanity acquittee for a period longer than he might have been imprisoned if convicted. But that is simply not the case here.
In particular circumstances, of course, it may be unconstitutional for a State to confine in a mental institution a person who is no longer insane. This would be a different case had Foucha challenged specific conditions of confinement — for instance, being forced to share a cell with an insane person, or being involuntarily treated after recovering his sanity. But Foucha has alleged nothing of the sort — all we know is that the State continues to confine him in a place called the Feliciana Forensic Facility. It is by no means clear that such confinement is invariably worse than, for example, confinement in a jail or other detention center — for all we know, an institution may provide a quieter, less violent atmosphere. I do not mean to suggest that that is the case — my point is only that the issue cannot be resolved in the abstract.