announced the judgment of the Court and delivered an opinion, in which The Chief Justice, Justice Stewart, and Justice Powell joined.

The question presented in this case is whether California’s “statutory rape” law, § 261.5 of the Cal. Penal Code Ann. (West Supp. 1981), violates the Equal Protection Clause of the Fourteenth Amendment. Section 261.5 defines unlawful sexual intercourse as “an act of sexual intercourse accomplished with a female not the wife of the perpetrator, where the female is under the age of 18 years.” The statute thus makes men alone criminally liable for the act of sexual intercourse.

In July 1978, a complaint was filed in the Municipal Court of Sonoma County, Cal., alleging that petitioner, then a 17%-year-old male, had had unlawful sexual intercourse with a female under the age of 18, in violation of § 261.5. The evidence adduced at a preliminary hearing showed that at approximately midnight on June 3, 1978, petitioner and two friends approached Sharon, a 16%-year-old female, and her sister as they waited at a bus stop. Petitioner and Sharon, *467who had already been drinking, moved away from the others and began to kiss. After being struck in the face for rebuffing petitioner’s initial advances, Sharon submitted to sexual intercourse with petitioner. Prior to trial, petitioner sought to set aside the information on both state and federal constitutional grounds, asserting that § 261.5 unlawfully discriminated on the basis of gender. The trial court and the California Court of Appeal denied petitioner’s request for relief and petitioner sought review in the Supreme Court of California.

The Supreme Court held that “section 261.5 discriminates on the basis of sex because only females may be victims, and only males may violate the section.” 25 Cal. 3d 608, 611, 601 P. 2d 572, 574. The court then subjected the classification to “strict scrutiny,” stating that it must be justified by a compelling state interest. It found that the classification was “supported not by mere social convention but by the immutable physiological fact that it is the female exclusively who can become pregnant.” Ibid. Canvassing “the tragic human costs of illegitimate teenage pregnancies,” including the large number of teenage abortions, the increased medical risk associated with teenage .pregnancies, and the social consequences of teenage childbearing, the court concluded that the State has a compelling interest in preventing such pregnancies. Because males alone can “physiologically cause the result which the law properly seeks to avoid,” the court further held that the gender classification was readily justified as a means of identifying offender and victim. For the reasons stated below, we affirm the judgment of the California Supreme Court.1

*468As is evident from our opinions, the Court has had some difficulty in agreeing upon the proper approach and analysis in cases involving challenges to gender-based classifications. The issues posed by such challenges range from issues of standing, see Orr v. Orr, 440 U. S. 268 (1979), to the appropriate standard of judicial review for the substantive classification. Unlike the California Supreme Court, we have not held that gender-based classifications are “inherently suspect” and thus we do not apply so-called “strict scrutiny” to those classifications. See Stanton v. Stanton, 421 U. S. 7 (1975). Our cases have held, however, that the traditional minimum rationality test takes on a somewhat “sharper focus” when gender-based classifications are challenged. See Craig v. Boren, 429 U. S. 190, 210 n." (1976) (Powell, J., concurring). In Reed v. Reed, 404 U. S. 71 (1971), for example, the Court stated that a gender-based classification will be upheld if it *469bears a “fair and substantial relationship” to legitimate state ends, while in Craig v. Boren, supra, at 197, the Court restated the test to require the classification to bear a “substantial relationship” to “important governmental objectives.”

Underlying these decisions is the principle that a legislature may not “make overbroad generalizations based on sex which are entirely unrelated to any differences between men and women or which demean the ability or social status of the affected class.” Parham v. Hughes, 441 U. S. 347, 354 (1979) (plurality opinion of Stewart, J.). But because the Equal Protection Clause does not “demand that a statute necessarily apply equally to all persons” or require “ 'things which are different in fact... to be treated in law as though they were the same/ ” Rinaldi v. Yeager, 384 U. S. 305, 309 (1966), quoting Tigner v. Texas, 310 U. S. 141, 147 (1940), this Court has consistently upheld statutes where the gender classification is not invidious, but rather realistically reflects the fact that the sexes are not similarly situated in certain circumstances. Parham v. Hughes, supra; Califano v. Webster, 430 U. S. 313 (1977); Schlesinger v. Ballard, 419 U. S. 498 (1975); Kahn v. Shevin, 416 U. S. 351 (1974). As the Court has stated, a legislature may “provide for the special problems of women.” Weinberger v. Wiesenfeld, 420 U. S. 636, 653 (1975).

Applying those principles to this case, the fact that the California Legislature criminalized the act of illicit sexual intercourse with a minor female is a sure indication of its intent or purpose to discourage that conduct.2 Precisely why the legislature desired that result is of course somewhat less clear. This Court has long recognized that “[¿Inquiries into congressional motives or purposes are a hazardous matter,” United States v. O’Brien, 391 U. S. 367, 383-384 (1968); Palmer v. Thompson, 403 U. S. 217, 224 (1971), and the *470search for the “actual” or “primary” purpose of a statute is likely to be elusive. Arlington Heights v. Metropolitan Housing Dev. Corp., 429 U. S. 252, 265 (1977); McGinnis v. Royster, 410 U. S. 263, 276-277 (1973). Here, for example, the individual legislators may have voted for the statute for a variety of reasons. Some legislators may have been concerned about preventing teenage pregnancies, others about protecting young females from physical injury or from the loss of “chastity,” and still others about promoting various religious and moral attitudes towards premarital sex.

The justification for the statute offered by the State, and accepted by the Supreme Court of California, is that the legislature sought to prevent illegitimate teenage pregnancies. That finding, of course, is entitled to great deference. Reitman v. Mulkey, 387 U. S. 369, 373-374 (1967). And although our cases establish that the State’s asserted reason for the enactment of a statute may be rejected, if it “could not have been a goal of the legislation,” Weinberger v. Wiesenfeld, supra, at 648, n. 16, this is not such a case.

We are satisfied not only that the prevention of illegitimate pregnancy is at least one of the “purposes” of the statute, but also that the State has a strong interest in preventing such pregnancy. At the risk of stating the obvious, teenage pregnancies, which have increased dramatically over the last two decades,3 have significant social, medical, and economic consequences for both the mother and her child, and the State.4 *471Of particular concern to the State is that approximately half of all teenage pregnancies end in abortion.5 And of those children who are born, their illegitimacy makes them likely candidates to become wards of the State.6

We need not be medical doctors to discern that young men and young women are not similarly situated with respect to the problems and the risks of sexual intercourse. Only women may become pregnant, and they suffer disproportionately the profound physical, emotional, and psychological consequences of sexual activity. The statute at issue here *472protects women from sexual intercourse at an age when those consequences are particularly severe.7

The question thus boils down to whether a State may attack the problem of sexual intercourse and teenage pregnancy directly by prohibiting a male from having sexual intercourse with a minor female.8 We hold that such a statute is *473sufficiently related to the State’s objectives to pass constitutional muster.

Because virtually all of the significant harmful and inescapably identifiable consequences of teenage pregnancy fall on the young female, a legislature acts well within its authority when it elects to punish only the participant who, by nature, suffers few of the consequences of his conduct. It is hardly unreasonable for a legislature acting to protect minor females to exclude them from punishment. Moreover, the risk of pregnancy itself constitutes a substantial deterrence to young females. No similar natural sanctions deter males. A criminal sanction imposed solely on males thus serves to roughly “equalize” the deterrents on the sexes.

We are unable to accept petitioner’s contention that the statute is impermissibly underinclusive and must, in order to pass judicial scrutiny, be broadened so as to hold the female as criminally liable as the male. It is argued that this statute is not necessary to deter teenage pregnancy because a gender-neutral statute, where both male and female would be subject to prosecution, would serve that goal equally well. The relevant inquiry, however, is not whether the statute is drawn as precisely as it might have been, but whether the line chosen by the California Legislature is within constitutional limitations. Kahn v. Shevin, 416 U. S., at 356, n. 10.

In any event, we cannot say that a gender-neutral statute would be as effective as the statute California has chosen to enact. The State persuasively contends that a gender-neutral statute would frustrate its interest in effective enforcement. Its view is that a female is surely less likely to report *474violations of the statute if she herself would be subject to criminal prosecution.9 In an area already fraught with prose-cutorial difficulties, we decline to hold that the Equal Protection Clause requires a legislature to enact a statute so broad that it may well be incapable of enforcement.10

*475We similarly reject petitioner’s argument that § 261.5 is impermissibly overbroad because it makes unlawful sexual intercourse with prepubescent females, who are, by definition, incapable of becoming pregnant. Quite apart from the fact that the statute could well be justified on the grounds that very young females are particularly susceptible to physical injury from sexual intercourse, see Rundlett v. Oliver, 607 F. 2d 495 (CA1 1979), it is ludicrous to suggest that the Constitution requires the California Legislature to limit the scope of its rape statute to older teenagers and exclude young girls.

There remains only petitioner’s contention that the statute is unconstitutional as it is applied to him because he, like Sharon, was under 18 at the time of sexual intercourse. Petitioner argues that the statute is flawed because it presumes that as between two persons under 18, the male is the culpable aggressor We find petitioner’s contentions unpersuasive. Contrary to his assertions, the statute does not rest on the assumption that males are generally the aggressors. It is instead an attempt by a legislature to prevent illegitimate teenage pregnancy by providing an additional deterrent for men. The age of the man is irrelevant since young men are as capable as older men of inflicting the harm sought to be. prevented.

In upholding the California statute we also recognize that this is not a case where a statute is being challenged on the grounds that it “invidiously discriminates” against females. *476To the contrary, the statute places a burden on males which is not shared by females. But we find nothing to suggest that men, because of past discrimination or peculiar disadvantages, are in need of the special solicitude of the courts. Nor is this a case where the gender classification is made “solely for . . . administrative convenience,” as in Frontiero v. Richardson, 411 U. S. 677, 690 (1973) (emphasis omitted), or rests on “the baggage of sexual stereotypes” as in Orr v. Orr, 440 U. S., at 283. As we have held, the statute instead reasonably reflects the fact that the consequences of sexual intercourse and pregnancy fall more heavily on the female than on the male.

Accordingly the judgment of the California Supreme Court is

Affirmed.

concurring.

Section 261.5, on its face, classifies on the basis of sex. A male who engages in sexual intercourse with an underage female who is not his wife violates the statute; a female who engages in sexual intercourse with an underage male who is not her husband does not.1 The petitioner contends that this state law, which punishes only males for the conduct in question, violates his Fourteenth Amendment right to the equal protection of the law. The Court today correctly rejects that contention.

A

At the outset, it should be noted that the statutory discrimination, when viewed as part of the wider scheme of California law, is not as clearcut as might at first appear. Females are not freed from criminal liability in California for engaging in sexual activity that may be harmful. It is unlawful, for example, for any person, of either sex, to molest, annoy, or contribute to the delinquency of anyone under 18 years of *477age.2 All persons are prohibited from committing “any lewd or lascivious act,” including consensual intercourse, with a child under 14.3 And members of both sexes may be convicted for engaging in deviant sexual acts with anyone under 18.4 Finally, females may be brought within the proscription of § 261.5 itself, since a female may be charged with aiding and abetting its violation.5

Section 261.5 is thus but one part of a broad statutory scheme that protects all minors from the problems and risks attendant upon adolescent sexual activity. To be sure, § 261.5 creates an additional measure of punishment for males who engage in sexual intercourse with females between the ages of 14 and 17.6 The question then is whether the Constitution prohibits a state legislature from imposing this additional sanction on a gender-specific basis.

B

The Constitution is violated when government, state or federal, invidiously classifies similarly situated people on the basis of the immutable characteristics with which they were

*478born. • Thus, detrimental racial classifications by government always violate the Constitution, for the simple reason that, so far as the Constitution is concerned, people of different races are always similarly situated. See Fullilove v. Klutznick, 448 U. S. 448, 522 (dissenting opinion); McLaughlin v. Florida, 379 U. S. 184, 198 (concurring opinion); Brown v. Board of Ed., 347 U. S. 483; Plessy v. Ferguson, 163 U. S. 537, 552 (dissenting opinion). By contrast, while detrimental gender classifications by government often violate the Constitution, they do not always do so, for the reason that there are differences between males and females that the Constitution necessarily recognizes. In this case we deal with the most basic of these differences: females can become pregnant as the result of sexual intercourse; males cannot.

As was recognized in Parham v. Hughes, 441 U. S. 347, 354, “a State is not free to make overbroad generalizations based on sex which are entirely unrelated to any differences between men and women or which demean the ability or social status of the affected class.” Gender-based classifications may not be based upon administrative convenience, or upon archaic assumptions about the proper roles of the sexes. Craig v. Boren, 429 U. S. 190; Frontiero v. Richardson, 411 U. S. 677; Reed v. Reed, 404 U. S. 71. But we have recognized that in certain narrow circumstances men and women are not similarly situated; in these circumstances a gender classification based on clear differences between the sexes is not invidious, and a legislative classification realistically based upon those differences is not unconstitutional. See Parham v. Hughes, supra; Califano v. Webster, 430 U. S. 313, 316-317; Schlesinger v. Ballard, 419 U. S. 498; cf. San Antonio Independent School Dist. v. Rodriguez, 411 U. S. 1, 59 (concurring opinion). “ [G] ender-based classifications are not invariably invalid. When men and women are not in fact similarly situated in the area covered by the legislation in question, the Equal Protection Clause is not violated.” Caban v. Mohammed, 441 U. S. 380, 398 (dissenting opinion).

*479Applying these principles to the classification enacted by the California Legislature, it is readily apparent that § 261.5 does not violate the Equal Protection Clause. Young women and men are not similarly situated with respect to the problems and risks associated with intercourse and pregnancy, and the statute is realistically related to the legitimate state purpose of reducing those problems and risks.

C

As the California Supreme Court’s catalog shows, the pregnant unmarried female confronts problems more numerous and more severe than any faced by her male partner.7 She alone endures the medical risks of pregnancy or abortion.8 She suffers disproportionately the social, educational, and emotional consequences of pregnancy.9 Recognizing this dis*480proportion, California has attempted to protect teenage females by prohibiting males from participating in the act necessary for conception.10

The fact that males and females are not-similarly situated with respect to the risks of sexual intercourse applies with the same force to males under 18 as it does to older males. The risk of pregnancy is a significant deterrent for unwed young females that is not shared by unmarried males, regardless of their age. Experienced observation confirms the commonsense notion that adolescent males disregard the possibility of pregnancy far more than do adolescent females.11 And to the extent that § 261.5 may punish males for intercourse with prepubescent females, that punishment is justifiable because of the substantial physical risks for prepubescent females that are not shared by their male counterparts.12

*481D

The petitioner argues that the California Legislature could have drafted the statute differently, so that its purpose would be accomplished more precisely. “But the issue, of course, is not whether the statute could have been drafted more wisely, but whether the lines chosen by the . . . [legislature are within constitutional limitations.” Kahn v. Shevin, 416 U. S. 351, 356, n. 10. That other States may have decided to attack the sanie problems more broadly, with gender-neutral statutes, does not mean that every State is constitutionally compelled to do so.13

E

In short, the Equal Protection Clause does not mean that the physiological differences between men and women must be disregarded. While those differences must never be permitted to become a pretext for invidious discrimination, no such discrimination is presented by this case. The Constitution surely does not require a State to pretend that demonstrable differences between men and women do not really exist.

concurring in the judgment.

It is gratifying that the plurality recognizes that “[a]t the risk of stating the obvious, teenage pregnancies . . . have increased dramatically over the last two decades” and “have significant social, medical, and economic consequences for both *482the mother and her child, and the State.” Ante, at 470 (footnotes omitted). There have been times when I have wondered whether the Court was capable of this perception, particularly when it has struggled with the different but not unrelated problems that attend abortion issues. See, for example, the opinions (and the dissenting opinions) in Beal v. Doe, 432 U. S. 438 (1977); Maher v. Roe, 432 U. S. 464 (1977) ; Poelker v. Doe, 432 U. S. 519 (1977); Harris v. McRae, 448 U. S. 297 (1980); Williams v. Zbaraz, 448 U. S. 358 (1980); and today’s opinion in H. L. v. Matheson, ante, p. 389.

Some might conclude that the two uses of the criminal sanction — here flatly to forbid intercourse in order to forestall teenage pregnancies, and in Matheson to prohibit a physician’s abortion procedure except upon notice to the parents of the pregnant minor — are vastly different proscriptions. But the basic social and privacy problems are much the same. Both Utah’s statute in Matheson and California’s statute in this case are legislatively created tools intended to achieve similar ends and addressed to the same societal concerns: the control and direction of young people’s sexual activities. The plurality opinion impliedly concedes as much when it notes that “approximately half of all teenage pregnancies end in abortion,” and that “those children who are born” are “likely candidates to become wards of the State,” Ante, at 471, and n. 6.

I, however, cannot vote to strike down the California statutory rape law, for I think it is a sufficiently reasoned and constitutional effort to control the problem at its inception. For me, there is an important difference between this state action and a State’s adamant and rigid refusal to face, or even to recognize, the “significant . . . consequences” — to the woman — of a forced or unwanted conception. I have found it difficult to rule constitutional, for example, state efforts to block, at that later point, a woman’s attempt to deal with the enormity of the problem confronting her, just as I have rejected state efforts to prevent women from rationally tak*483ing steps to prevent that problem from arising. See, e. g., Carey v. Population Services International, 431 U. S. 678 (1977). See also Griswold v. Connecticut, 381 U. S. 479 (1966). In contrast, I am persuaded that, although a minor has substantial privacy rights in intimate affairs connected with procreation, California’s efforts to prevent teenage pregnancy are to be viewed differently from Utah’s efforts to inhibit a woman from dealing with pregnancy once it has become an inevitability.

Craig v. Boren, 429 U. S. 190 (1976), was an opinion which, in large part, I joined, id., at 214. The plurality opinion in the present case points out, ante, at 468-469, the Court’s respective phrasings of the applicable test in Reed v. Reed, 404 U. S. 71, 76 (1971), and in Craig v. Boren, 429 U. S., at 197. I vote to affirm the judgment of the Supreme Court of California and to uphold the State’s gender-based classification on that test and as exemplified by those two cases and, by Schlesinger v. Ballard, 419 U. S. 498 (1975); Weinberger v. Wiesenfeld, 420 U. S. 636 (1975); and Kahn v. Shevin, 416 U. S. 351 (1974).

I note, also, that § 261.5 of the California Penal Code is just one of several California statutes intended to protect the juvenile. Justice Stewart, in his concurring opinion, appropriately observes that § 261.5 is “but one part of a broad statutory scheme that protects all minors from the problems and risks attendant upon adolescent sexual activity.” Ante, at 477.

I think, too, that it is only fair, with respect to this particular petitioner, to point out that his partner, Sharon, appears not to have been an unwilling participant in at least the initial stages of the intimacies that took place the night of June 3, 1978.* Petitioner’s and Sharon’s nonacquaintance *484with each other before the incident; their drinking; their withdrawal from the others of the group; their foreplay, in which she willingly participated and seems to have encour*485aged; and the closeness of their ages (a difference of only one year and 18 days) are factors that should make this case an unattractive one to prosecute at all, and especially to pros*486ecute as a felony, rather than as a misdemeanor chargeable under § 261.5. But the State has chosen to prosecute in that *487manner, and the facts, I reluctantly conclude, may fit the crime.

with whom Justices White and Marshall join, dissenting.

I

It is disturbing to find the Court so splintered on a case that presents such a straightforward issue: Whether the admittedly gender-based classification in Cal. Penal Code Ann. §261.5 (West Supp. 1981) bears a sufficient relationship to the State’s asserted goal of preventing teenage pregnancies to survive the “mid-level” constitutional scrutiny mandated by Craig v. Boren, 429 U. S. 190 (1976).1 Applying the analytical framework provided by our precedents, I am convinced that there is only one proper resolution of this issue: the classification must be declared unconstitutional. I fear that the plurality opinion and Justices Stewart and Blackmun reach the opposite result by placing too much emphasis on the desirability of achieving the State’s asserted statutory goal — prevention of teenage pregnancy — and not enough emphasis on the fundamental question of whether the sex-based discrim*489ination in the California statute is substantially related to the achievement of that goal.2

II

After some uncertainty as to the proper framework for analyzing equal protection challenges to statutes containing gender-based classifications, see ante, at 468, this Court settled upon the proposition that a statute containing a gender-based classification cannot withstand constitutional challenge unless *490the classification is substantially related to the achievement of an important governmental objective. Kirchberg v. Feenstra, ante, at 459; Wengler v. Druggists Mutual Ins. Co., 446 U. S. 142, 150 (1980); Califano v. Westcott, 443 U. S. 76, 85 (1979); Caban v. Mohammed, 441 U. S. 380, 388 (1979); Orr v. Orr, 440 U. S. 268, 279 (1979); Califano v. Goldfarb, 430 U. S. 199, 210-211 (1977); Califano v. Webster, 430 U. S. 313, 316-317 (1977); Craig v. Boren, supra, at 197. This analysis applies whether the classification discriminates against males or against females. Caban v. Mohammed, supra, at 394; Orr v. Orr, supra, at 278-279; Craig v. Boren, supra, at 204. The burden is on the government to prove both the importance of its asserted objective and the substantial relationship between the classification and that objective. See Kirchberg v. Feenstra, ante, at 461; Wengler v. Druggists Mutual Ins. Co., supra, at 151-152; Caban v. Mohammed, supra, at 393; Craig v. Boren, supra, at 204. And the State cannot meet that burden without showing that a gender-neutral statute would be a less effective means of achieving that goal. Wengler v. Druggists Mutual Ins. Co., supra, at 151-152; Orr v. Orr, supra, at 281, 283.3

The State of California vigorously asserts that the “important governmental objective” to be served by § 261.5 is the prevention of teenage pregnancy. It claims that its statute furthers this goal by deterring sexual activity by males — the class of persons it considers more responsible for causing those pregnancies.4 But even assuming that prevention of teenage *491pregnancy is an important governmental objective and that it is in fact an objective of § 261.5, see infra, at 494-496, California still has the burden of proving that there are fewer teenage pregnancies under its gender-based statutory rape law than there would be if the law were gender neutral. To meet this burden, the State must show that because its statutory rape law punishes only males, and not females, it more effectively deters minor females from having sexual intercourse.5

The plurality assumes that a gender-neutral statute would be less effective than § 261.5 in deterring sexual activity because a gender-neutral statute would create significant enforcement problems. The plurality thus accepts the State’s assertion that

“a female is surely less likely to report violations of the statute if she herself would be subject to criminal prose*492cution. In an area already fraught with prosecutorial difficulties, we decline to hold that the Equal Protection Clause requires a legislature to enact a statute so broad that it may well be incapable of enforcement.” Ante, at 473-474 (footnotes omitted).

However, a State’s bare assertion that its gender-based statutory classification substantially furthers an important governmental interest is not enough to meet its burden of proof under Craig v. Boren. Rather, the State must produce evidence that will persuade the court that its assertion is true. See Craig v. Boren, 429 U. S., at 200-204.

The State has not produced such evidence in this case. Moreover, there are at least two serious flaws in the State’s assertion that law enforcement problems created by a gender-neutral statutory rape law would make such a statute less effective than a gender-based statute in deterring sexual activity.

First, the experience of other jurisdictions, and California itself, belies the plurality’s conclusion that a gender-neutral statutory rape law “may well be incapable of enforcement.” There are now at least 37 States that have enacted gender-neutral statutory rape laws. Although most of these laws protect young persons (of either sex) from the sexual exploitation of older individuals, the laws of Arizona, Florida, and Illinois permit prosecution of both minor females and minor males for engaging in mutual sexual conduct.6 California has introduced no evidence that those States have been handi*493capped by the enforcement problems the plurality finds so persuasive.7 Surely, if those States could provide such evidence, we might expect that California would have introduced it.

In addition, the California Legislature in recent years has revised other sections of the Penal Code to make them gender-neutral. For example, Cal. Penal Code Ann. §§286 (b)(1) and 288a (b)(1) (West Supp. 1981), prohibiting sodomy and oral copulation with a “person who is under 18 years of age,” could cause two minor homosexuals to be subjected to criminal sanctions for engaging in mutually consensual conduct. Again, the State has introduced no evidence to explain why a gender-neutral statutory rape law would be any more difficult to enforce than those statutes.

The second flaw in the State’s assertion is that even assuming that a gender-neutral statute would be more difficult to enforce, the State has still not shown that those enforcement problems would make such a statute less effective than a gender-based statute in deterring minor females from engaging in sexual intercourse.8 Common sense, however, suggests *494that a gender-neutral statutory rape law is potentially a greater deterrent of sexual activity than a gender-based law, for the simple reason that a gender-neutral law subjects both men and women to criminal sanctions and thus arguably has a deterrent effect on twice as many potential violators. Even if fewer persons were prosecuted under the gender-neutral law, as the State suggests, it would still be true that twice as many persons would be subject to arrest. The State’s failure to prove that a gender-neutral law would be a less effective deterrent than a gender-based law, like the State’s failure to prove that a gender-neutral law would be difficult to enforce, should have led this Court to invalidate § 261.5.

Ill

Until very recently, no California court or commentator had suggested that the purpose of California’s statutory rape law was to protect young women from the risk of pregnancy. Indeed, the historical development of § 261.5 demonstrates that the law was initially enacted on the premise that young women, in contrast to young men, were to be deemed legally incapable of consenting to an act of sexual intercourse.9 Be*495cause their chastity was considered particularly precious, those young women were felt to be uniquely in need of the State’s protection.10 In contrast, young men were assumed to *496be capable of making such decisions for themselves; the law therefore did not offer them any special protection.

It is perhaps because the gender classification in California’s statutory rape law was initially designed to further these outmoded sexual stereotypes, rather than to reduce the incidence of teenage pregnancies, that the State has been unable to demonstrate a substantial relationship beween the classification and its newly asserted goal. Cf. Califano v. Goldfarb, 430 U. S., at 223 (Stevens, J., concurring in judgment). But whatever the reason, the State has not shown that Cal. Penal Code § 261.5 is any more effective than a gender-neutral law would be in deterring minor females from engaging in sexual intercourse. It has therefore not met its burden of proving that the statutory classification is substantially related to the achievement of its asserted goal.

I would hold that § 261.5 violates the Equal Protection Clause of the Fourteenth Amendment, and I would reverse the judgment of the California Supreme Court.

dissenting.

Local custom and belief — rather than statutory laws of venerable but doubtful ancestry — will determine the volume of sexual activity among unmarried teenagers.1 The empiri*497cal evidence cited by the plurality demonstrates the futility of the notion that a statutory prohibition will significantly affect the volume of that activity or provide a meaningful solution to the problems created by it.2 Nevertheless, as a matter of constitutional power, unlike my Brother Brennan, see ante, at 491, n. 5, I would have no doubt about the validity of a state law prohibiting all unmarried teenagers from engaging in sexual intercourse. The societal interests in reducing the incidence of venereal disease and teenage pregnancy are sufficient, in my judgment, to justify a prohibition of conduct that increases the risk of those harms.3

My conclusion that a nondiscriminatory prohibition would be constitutional does not help me answer the question whether a prohibition applicable to only half of the joint participants in the risk-creating conduct is also valid. It cannot be true that the validity of a total ban is an adequate justification for a selective prohibition; otherwise, the constitutional objection to discriminatory rules would be meaningless. The question in this case is whether the difference between males and females justifies this statutory discrimination based entirely on sex.4

*498The fact that the Court did not immediately acknowledge that the capacity to become pregnant is what primarily differentiates the female from the male5 does not impeach the validity of the plurality’s newly found wisdom. I think the plurality is quite correct in making the assumption that the joint act that this law seeks to prohibit creates a greater risk of harm for the female than for the male. But the plurality surely cannot believe that the risk of pregnancy confronted by the female — any more than the risk of venereal disease confronted by males as well as females — has provided an effective deterrent to voluntary female participation in the risk-creating conduct. Yet the plurality’s decision seems to rest on the assumption that the California Legislature acted on the basis of that rather fanciful notion.

*499In my judgment, the fact that a class of persons is especially vulnerable to a risk that a statute is designed to avoid is a reason for making the statute applicable to that class. The argument that a special need for protection provides a rational explanation for an exemption is one I simply do not comprehend.6

In this case, the fact that a female confronts a greater risk of harm than a male is a reason for applying the prohibition to her — not a reason for granting her a license to use her own judgment on whether or not to assume the risk. Surely, if we examine the problem from the point of view of society’s interest in preventing the risk-creating conduct from occurring at all, it is irrational to exempt 50% of the potential violators. See dissent of Justice Brennan, ante, at 493-494. And, if we view the government’s interest as that of a parens patriae seeking to protect its subjects from harming themselves, the discrimination is actually perverse. Would a rational parent making rules for the conduct of twin children of opposite sex simultaneously forbid the son and authorize the daughter to engage in conduct that is especially harmful to the daughter? That is the effect of this statutory classification.

If pregnancy or some other special harm is suffered by one of the two participants in the prohibited act, that special harm no doubt would constitute a legitimate mitigating factor in deciding what, if any, punishment might be appropriate in a given case. But from the standpoint of fashioning a general preventive rule — or, indeed, in determining appropriate punishment when neither party in fact has suffered any spe*500cial harm — I regard a total exemption for the members of the more endangered class as utterly irrational.

In my opinion, the only acceptable justification for a general rule requiring disparate treatment of the two participants in a joint act must be a legislative judgment that one is more guilty than the other. The risk-creating conduct that this statute is designed to prevent requires the participation of two persons — one male and one female.7 In many situations it is probably true that one is the aggressor and the other is either an unwilling, or at least a less willing, participant in the joint act. If a statute authorized punishment of only one participant and required the prosecutor to prove that that participant had been the aggressor, I assume that the discrimination would be valid. Although the question is less clear, I also assume, for the purpose of deciding this case, that it would be permissible to punish only the male participant, if one element of the offense were proof that he had been the aggressor, or at least in some respects the more responsible participant in the joint act. The statute at issue in this case, however, requires no such proof. The question raised by this statute is whether the State, consistently with the Federal Constitution, may always punish the male and never the female when they are equally responsible or when the female is the more responsible of the two.

It would seem to me that an impartial lawmaker could give only one answer to that question. The fact that the California Legislature has decided to apply its prohibition only to *501the male may reflect a legislative judgment that in the typical case the male is actually the more guilty party. Any such judgment must, in turn, assume that the decision to engage in the risk-creating conduct is always — or at least typically— a male decision. If that assumption is valid, the statutory classification should also be valid. But what is the support for the assumption? It is not contained in the record of this case or in any legislative history or scholarly st,udy that has been called to our attention. I think it is supported to some extent by traditional attitudes toward male-female relationships. But the possibility that such a habitual attitude may reflect nothing more than an irrational prejudice makes it an insufficient justification for discriminatory treatment that is otherwise blatantly unfair. For, as I read this statute, it requires that one, and only one, of two equally guilty wrongdoers be stigmatized by a criminal conviction.

I cannot accept the State’s argument that the constitutionality of the discriminatory rule can be saved by an assumption that prosecutors will commonly invoke this statute only in cases that actually involve a forcible rape, but one that cannot be established by proof beyond a reasonable doubt.8 That assumption implies that a State has a legitimate interest in convicting a defendant on evidence that is constitutionally insufficient. Of course, the State may create a lesser-included offense that would authorize punishment of the more guilty party, but surely the interest in obtaining convictions on in*502adequate proof cannot justify a statute that punishes one who is equally or less guilty than his partner.9

Nor do I find at all persuasive the suggestion that this discrimination is adequately justified by the desire to encourage females to inform against their male partners. Even if the concept of a wholesale informant’s exemption were an acceptable enforcement device, what is the justification for defining the exempt class entirely by reference to sex rather than by reference to a more neutral criterion such as relative innocence? Indeed, if the exempt class is to be composed entirely of members of one sex, what is there to support the view that the statutory purpose will be better served by granting the informing license to females rather than to males? If a discarded male partner informs on a promiscuous female, a timely threat of prosecution might well prevent the precise harm the statute is intended to minimize.

Finally, even if my logic is faulty and there actually is some speculative basis for treating equally guilty males and females differently, I still believe that any such speculative justification would be outweighed by the paramount interest in evenhanded enforcement of the law. A rule that authorizes punishment of only one of two equally guilty wrongdoers violates the essence of the constitutional requirement that the sovereign must govern impartially.

I respectfully dissent.