delivered the opinion of the Court.

Appellee Dianne Rawlinson sought employment with the Alabama Board of Corrections as a prison guard, called in Alabama a "correctional counselor.” After her application was rejected, she brought this class suit under Title YII of the Civil Rights Act of 1964, 78 Stat. 253, as amended, 42 U. S. C. § 2000e et seq. (1970 ed. and Supp. V), and under 42 U. S. C. § 1983, alleging that she had been denied employment because of her sex in violation of federal law. A three-judge Federal District Court for the Middle District of Alabama decided in her favor. Mieth v. Dothard, 418 F. Supp. 1169. We noted probable jurisdiction of this appeal from the District Court’s judgment. 429 U. S. 976.1

I

At the time she applied for a position as correctional counselor trainee, Rawlinson was a 22-year-old college graduate whose major course of study had been correctional psychology. She was refused employment because she failed to meet the minimum 120-pound weight requirement estab*324lished by an Alabama statute. The statute also establishes a height minimum of 5 feet 2 inches.2

After her application was rejected because of her weight, Rawlinson filed a charge with the Equal Employment Opportunity Commission, and ultimately received a right-to-sue letter.3 She then filed a complaint in the District Court on behalf of herself and other similarly situated women, challenging the statutory height and weight minima as violative of Title VII and the Equal Protection Clause of the Fourteenth Amendment.4 A three-judge court was convened.5 While the suit was pending, the Alabama Board of Correc*325tions adopted Administrative Regulation 204, establishing gender criteria for assigning correctional counselors to maximum-security institutions for “contact positions,” that is, positions requiring continual close physical proximity to inmates of the institution.6 Rawlinson amended her class-action *326complaint by adding a challenge to Regulation 204 as also violative of Title VII and the Fourteenth Amendment.

Like most correctional facilities in the United States,7 Alabama’s prisons are segregated on the basis of sex. Currently the Alabama Board of Corrections operates four major all-male penitentiaries — Holman Prison, Kilby Corrections Facility, G. K. Fountain Correction Center, and Draper Correctional Center. The Board also operates the Julia Tutwiler Prison for Women, the Frank Lee Youth Center, the Number Four Honor Camp, the State Cattle Ranch, and nine Work Release Centers, one of which is for women. The Julia Tutwiler Prison for Women and the four male penitentiaries are maximum-security institutions. Their inmate living quarters are for the most part large dormitories, with communal showers and toilets that are open to the dormitories and hallways. The Draper and Fountain penitentiaries carry on extensive farming operations, making necessary a large number of strip searches for contraband when prisoners re-enter the prison buildings.

A correctional counselor’s primary duty within these institutions is to maintain security and control of the inmates *327by continually supervising and observing their activities.8 To be eligible for consideration as a correctional counselor, an applicant must possess a valid Alabama driver’s license, have a high school education or its equivalent, be free from physical defects, be between the ages of 20% years and 45 years at the time of appointment, and fall between the minimum height and weight requirements of 5 feet 2 inches, and 120 pounds, and the maximum of 6 feet 10 inches, and 300 pounds. Appointment is by merit, with a grade assigned each applicant based on experience and education. No written examination is given.

At the time this litigation was in the District Court, the Board of Corrections employed a total of 435 people in various correctional counselor positions, 56 of whom were women. Of those 56 women, 21 were employed at the Julia Tutwiler Prison for Women, 13 were employed in noncontact positions at the four male maximum-security institutions, and the remaining 22 were employed at the other institutions operated by the Alabama Board of Corrections. Because most of Alabama’s prisoners are held at the four maximum-security male penitentiaries, 336 of the 435 correctional counselor jobs were in those institutions, a majority of them concededly in the “contact” classification.9 Thus, even though meeting the statutory height and weight requirements, women applicants could under Regulation 204 com*328pete equally with men for only about 25% of the correctional counselor jobs available in the Alabama prison system.

II

In enacting Title VII, Congress required “the removal of artificial, arbitrary, and unnecessary barriers to employment when the barriers operate invidiously to discriminate on the basis of racial or other impermissible classification.” Griggs v. Duke Power Co., 401 U. S. 424, 431. The District Court found that the minimum statutory height and weight requirements that applicants for employment as correctional counselors must meet constitute the sort of arbitrary barrier to equal employment opportunity that Title VII forbids.10 The appellants assert that the District Court erred both in finding that the height and weight standards discriminate against women, and in its refusal to find that, even if they do, these standards are justified as “job related.”

A

The gist of the claim that the statutory height and weight requirements discriminate against women does not involve an assertion of purposeful discriminatory motive.11 It is as*329serted, rather, that these facially neutral qualification standards work in fact disproportionately to exclude women from eligibility for employment by the Alabama Board of Corrections. We dealt in Griggs v. Duke Power Co., supra, and Albemarle Paper Co. v. Moody, 422 U. S. 405, with similar allegations that facially neutral employment standards disproportionately excluded Negroes from employment, and those cases guide our approach here.

Those cases make clear that to establish a prima facie case of discrimination, a plaintiff need only show that the facially neutral standards in question select applicants for hire in a significantly discriminatory pattern. Once it is thus shown that the employment standards are discriminatory in effect, the employer must meet “the burden of showing that any given requirement [has] ... a manifest relationship to the employment in question.” Griggs v. Duke Power Co., supra, at 432. If the employer proves that the challenged requirements are job related, the plaintiff may then show that other selection devices without a similar discriminatory effect would also “serve the employer’s legitimate interest in ‘efficient and trustworthy workmanship.’ ” Albemarle Paper Co. v. Moody, supra, at 425, quoting McDonnell Douglas Corp. v. Green, 411 U. S. 792, 801.

Although women 14 years of age or older compose 52.75% of the Alabama population and 36.89% of its total labor force, they hold only 12.9% of its correctional counselor positions. In considering the effect of the minimum height and weight standards on this disparity in rate of hiring between the sexes, the District Court found that the 5'2"-requirement would operate to exclude 33.29% of the women in the United States between the ages of 18-79, while excluding only 1.28% of men between the same ages. The 120-pound weight restriction would exclude 22.29% of the women and 2.35% of the men in this age group. When the height and weight restrictions are combined, Alabama’s statutory standards would exclude 41.13% of the female popula*330tion while excluding less than 1% of the male population.12 Accordingly, the District Court found that Rawlinson had made out a prima facie case of unlawful sex discrimination.

The appellants argue that a showing of disproportionate impact on women based on generalized national statistics should not suffice to establish a prima facie case. They point in particular to Rawlinson’s failure to adduce comparative statistics concerning actual applicants for correctional counselor positions in Alabama. There is no requirement, however, that a statistical showing of disproportionate impact must always be based on analysis of the characteristics of actual applicants. See Griggs v. Duke Power Co., supra, at 430. The application process itself might not adequately reflect the actual potential applicant pool, since otherwise qualified people might be discouraged from applying because of a self-recognized inability to meet the very standards challenged as being discriminatory. See Teamsters v. United States, 431 U. S. 324, 365-367. A potential applicant could easily determine her height and weight and conclude that to make an application would be futile. Moreover, reliance on general population demographic data was not misplaced where there was no reason to suppose that physical height and weight characteristics of Alabama men and women differ markedly from those of the national population.

*331For these reasons, we cannot say that the District Court was wrong in holding that the statutory height and weight standards had a discriminatory impact on women applicants. The plaintiffs in a case such as this are not required to exhaust every possible source of evidence, if the evidence actually presented on its face conspicuously demonstrates a job requirement’s grossly discriminatory impact. If the employer discerns fallacies or deficiencies in the data offered by the plaintiff, he is free to adduce countervailing evidence of his own. In this case no such effort was made.13

B

We turn, therefore, to the appellants’ argument that they have rebutted the prima facie case of discrimination by showing that the height and weight requirements are job related. These requirements, they say, have a relationship to strength, a sufficient but unspecified amount of which is essential to effective job performance as a correctional counselor. In the District Court, however, the appellants produced no evidence correlating the height and weight requirements with the requisite amount of strength thought essential to good job performance. Indeed, they failed to offer evidence of any kind in specific justification of the statutory standards.14

*332If the job-related quality that the appellants identify is bona fide, their purpose could be achieved by adopting and validating a test for applicants that measures strength directly.15 Such a test, fairly administered, would fully satisfy the standards of Title VII because it would be one that “measure [s] the person for the job and not the person in the abstract.” Griggs v. Duke Power Co., 401 U. S., at 436. But nothing in the present record even approaches such a measurement.

For the reasons we have discussed, the District Court was not in error in holding that Title VII of the Civil Rights Act of 1964, as amended, prohibits application of the statutory height and weight requirements to Rawlinson and the class she represents.

Ill

Unlike the statutory height and weight requirements, Regulation 204 explicitly discriminates against women on the basis of their sex.16 In defense of this overt discrimination, *333the appellants rely on § 703 (e) of Title VII, 42 U. S. C. §2000e-2(e), which permits sex-based discrimination “in those certain instances where . . . sex ... is a bona fide occupational qualification reasonably necessary to the normal operation of that particular business or enterprise.”

The District Court rejected the bona-fide-occupational-qualification (bfoq) defense, relying on the virtually uniform view of the federal courts that § 703 (e) provides only the narrowest of exceptions to the general rule requiring equality of employment opportunities. This view has been variously formulated. In Diaz v. Pan American World Airways, 442 F. 2d 385, 388, the Court of Appeals for the Fifth Circuit held that “discrimination based on sex is valid only when the essence of the business operation would be undermined by not hiring members of one sex exclusively.” (Emphasis in original.) In an earlier case, Weeks v. Southern Bell Tel. & Tel. Co., 408 F. 2d 228, 235, the same court said that an employer could rely on the bfoq exception only by proving “that he had reasonable cause to believe, that is, a factual basis for believing, that all or substantially all women would be unable to perform safely and efficiently the duties of the job involved.” See also Phillips v. Martin Marietta Corp., 400 U. S. 542. But whatever the verbal formulation, the federal courts have agreed that it is impermissible under Title VII to refuse to hire an individual woman or man on the basis of stereotyped characterizations of the sexes,17 and the District *334Court in the present case held in effect that Regulation 204 is based on just such stereotypical assumptions.

We are persuaded — by the restrictive language of § 703 (e), the relevant legislative history,18 and the consistent interpretation of the Equal Employment Opportunity Commission 19 — that the bfoq exception was in fact meant to be an extremely narrow exception to the general prohibition of discrimination on the basis of sex.20 In the particular factual circumstances of this case, however, ,we conclude that the District Court erred in rejecting the State’s contention that Regulation 204 falls within the narrow ambit of the bfoq exception.

The environment in Alabama’s penitentiaries is a peculiarly inhospitable one for human beings of whatever sex. Indeed, a Federal District Court has held that the conditions of confinement in the prisons of the State, characterized by “rampant violence” and a “jungle atmosphere,” are constitutionally intolerable. Pugh v. Locke, 406 F. Supp. 318, 325 (MD Ala.). The record in the present case shows that *335because of inadequate staff and facilities, no attempt is made in the four maximum-security male penitentiaries to classify or segregate" inmates according to their offense or level of dangerousness — a procedure that, according to expert testimony, is essential to effective penological administration. Consequently, the estimated 20% of the male prisoners who are sex offenders are scattered throughout the penitentiaries' dormitory facilities.

In this environment of violence and disorganization, it would be an oversimplification to characterize Regulation 204 as an exercise in “romantic paternalism.” Cf. Frontiero v. Richardson, 411 U. S. 677, 684. In the usual case, the argument that a particular job is too dangerous for women may appropriately be met by the rejoinder that it is the purpose of Title VII to allow the individual woman to make that choice for herself.21 More is at stake in this case, however, than an individual woman’s decision to weigh and accept the risks of employment in a “contact” position in a maximum-security male prison.

The essence of a correctional counselor’s job is to maintain prison security. A woman’s relative ability to maintain order in a male, maximum-security, unclassified penitentiary of the type Alabama now runs could be directly reduced by her womanhood. There is a basis in fact for expecting that sex offenders who have criminally assaulted women in the past would be moved to do so again if access to women were established within the prison. There would also be a real risk that other inmates, deprived of a normal heterosexual environment, would assault women guards because they were women.22 In a prison system where violence is the order *336of the day, where inmate access to guards is facilitated by dormitory living arrangements, where every institution is understaffed, and where a substantial portion of the inmate population is composed of sex offenders mixed at random with other prisoners, there are few visible deterrents to inmate assaults on women custodians.

Appellee Rawlinson’s own expert testified that dormitory housing for aggressive inmates poses a greater security problem than single-cell lockups, and further testified that it would be unwise to use women as guards in a prison where even 10% of the inmates had been convicted of sex crimes and were not segregated from the other prisoners.23 The likelihood that inmates would assault a woman because she was a woman would pose a real threat not only to the victim of the assault but also to the basic control of the penitentiary and protection of its inmates and the other security personnel. The employee’s very womanhood would thus directly undermine her capacity to provide the security that is the essence of a correctional counselor’s responsibility.

There was substantial testimony from experts on both sides of this litigation that the use of women as guards in "contact” positions under the existing conditions in Alabama maximum-security male penitentiaries would pose a substantial security problem, directly linked to the sex of the prison guard. On the basis of that evidence, we conclude that the District Court was in error in ruling that being male is not a bona fide occupational qualification for the job of *337correctional counselor in a “contact” position in an Alabama male maximum-security penitentiary.24

The judgment is accordingly affirmed in part and reversed in part, and the case is remanded to the District Court for further proceedings consistent with this opinion.

It is so ordered.

concurring in the result and concurring in part.

I agree with, and join, Parts I and III of the Court’s opinion in this case and with its judgment. While I also agree with the Court’s conclusion in Part II of its opinion, holding that the District Court was “not in error” in holding the statutory height and weight requirements in this case to be invalidated by Title VII, ante, at 332, the issues with which that Part deals are bound to arise so frequently that I feel obliged to separately state the reasons for my agreement with its result. I view affirmance of the District Court in this respect as essentially dictated by the peculiarly limited factual and legal justifications offered below by appellants on behalf of the statutory requirements. For that reason-, I do not believe — and do not read the Court’s opinion as holding — that all or even many of the height and weight requirements imposed by States on applicants for a multitude of law enforcement agency jobs are pretermit-ted by today’s decision.

I agree that the statistics relied upon in this case are sufficient, absent rebuttal, to sustain a finding of a prima *338facie violation of § 703 (a)(2), in that they reveal a significant discrepancy between the numbers of men, as opposed to women, who are automatically disqualified by reason of the height and weight requirements. The fact that these statistics are national figures of height and weight, as opposed to statewide or pool-of-labor-force statistics, does not seem to me to require us to hold that the District Court erred as a matter of law in admitting them into evidence. See Hamling v. United States, 418 U. S. 87, 108, 124-125 (1974); cf. Zenith Corp. v. Hazeltine, 395 U. S. 100, 123-125 (1969). It is for the District Court, in the first instance, to determine whether these statistics appear sufficiently probative of the ultimate fact in issue — whether a given job qualification requirement has a disparate impact on some group protected by Title VII. Hazelwood School Dist. v. United States, ante, at 312-313; see Hamling v. United States, supra, at 108, 124-125; Mayor v. Educational Equality League, 415 U. S. 605, 621 n. 20 (1974); see also McAllister v. United States, 348 U. S. 19 (1954); United States v. Yellow Cab Co., 338 U. S. 338, 340-342 (1949). In making this determination, such statistics are to be considered in light of all other relevant facts and circumstances. Cf. Teamsters v. United States, 431 U. S. 324, 340 (1977). The statistics relied on here do not suffer from the obvious lack of relevancy of the statistics relied on by the District Court in Hazelwood School Dist. v. United States, ante, at 308. A reviewing court cannot say as a matter of law that they are irrelevant to the contested issue or so lacking in reliability as to be inadmissible.

If the defendants in a Title VII suit believe there to be any reason to discredit plaintiffs’ statistics that does not appear on their face, the opportunity to challenge them is available to the defendants just as in any other lawsuit. They may endeavor to impeach the reliability of the statistical evidence, they may offer rebutting evidence, or they may disparage in arguments or in briefs the probative weight which *339the plaintiffs’ evidence should be accorded. Since I agree with the Court that appellants made virtually no such effort, ante, at 331,1 also agree with it that the District Court cannot be said to have erred as a matter of law in finding that a prima facie case had been made out in the instant case.

While the District Court’s conclusion is by no means required by the proffered evidence, I am unable to conclude that the District Court’s finding in that respect was clearly erroneous. In other cases there could be different evidence which could lead a district court to conclude that height and weight are in fact an accurate enough predictor of strength to justify, under all the circumstances, such minima. Should the height and weight requirements be found to advance the job-related qualification of strength sufficiently to rebut the prima facie case, then, under our cases, the burden would shift back to appellee Rawlinson to demonstrate that other tests, without such disparate effect, would also meet that concern. Albemarle Paper Co. v. Moody, 422 U. S. 405, 425 (1975). But, here, the District Court permissibly concluded that appellants had not shown enough of a nexus even to rebut the inference.

Appellants, in order to rebut the prima facie case under the statute, had the burden placed on them to advance job-related reasons for the qualification. McDonnell Douglas Corp. v. Creen, 411 U. S. 792, 802 (1973). This burden could be shouldered by offering evidence or by making legal arguments not dependent on any new evidence. The District Court was confronted, however, with only one suggested job-related reason for the qualification — that of strength. Appellants argued only the job-relatedness of actual physical strength; they did not urge that an equally job-related qualification for prison guards is the appearance of strength. As the Court notes, the primary job of correctional counselor in Alabama prisons “is to maintain security and control of the inmates . . .,” ante, at 326, a function that I at least would *340imagine is aided by the psychological impact on prisoners of the presence of tall and heavy guards. If the appearance of strength had been urged upon the District Court here as a reason for the height and weight minima, I think that the District Court would surely have been entitled to reach a different result than it did. For, even if not perfectly correlated, I would think that Title VII would not preclude a State from saying that anyone under 5'2" or 120 pounds, no matter how strong in fact, does not have a sufficient appearance of strength to be a prison guard.

But once the burden has been placed on the defendant, it is then up to the defendant to articulate the asserted job-related reasons underlying the use of the minima. McDonnell Douglas Corp. v. Green, supra., at 802; Griggs v. Duke Power Co., 401 U. S. 424, 431 (1971); Albemarle Paper Co. v. Moody, supra, at 425. Because of this burden, a reviewing court is not ordinarily justified in relying on arguments in favor of a job qualification that were not first presented to the trial court. Cf. United States v. Arnold, Schwinn & Co., 388 U. S. 365, 374 n. 5 (1967); Thomas v. Taylor, 224 U. S. 73, 84 (1912); Bell v. Bruen, 1 How. 169, 187 (1843). As appellants did not even present the “appearance of strength” contention to the District Court as an asserted job-related reason for the qualification requirements, I agree that their burden was not met. The District Court's holding thus did not deal with the question of whether such an assertion could or did rebut appellee Rawlinson’s prima facie case.

with whom Mr. Justice Brennan joins, concurring in part and dissenting in part.

I agree entirely with the Court’s analysis of Alabama’s height and weight requirements for prison guards, and with its finding that these restrictions discriminate on the basis of sex in violation of Title VII. Accordingly, I join Parts I and II of the Court’s opinion. I also agree with much of the Court’s general discussion in Part III of the bona-fide-occupational-*341qualification exception contained in § 703 (e) of Title VII.1 The Court is unquestionably correct when it holds “that the bfoq exception was in fact meant to be an extremely narrow exception to the general prohibition of discrimination on the basis of sex.” Ante, at 334. See Phillips v. Martin Marietta Corp., 400 U. S. 542, 544 (1971) (Marshall, J., concurring). I must, however, respectfully disagree with the Court’s application of the bfoq exception in this case.

The Court properly rejects two proffered justifications for denying women jobs as prison guards. It is simply irrelevant here that a guard’s occupation is dangerous and that some women might be unable to protect themselves adequately. Those themes permeate the testimony of the state officials below, but as the Court holds, “the argument that a particular job is too dangerous for women” is refuted by the “purpose of Title VII to allow the individual woman to make that choice for herself.” Ante, at 335. Some women, like some men, undoubtedly are not qualified and do not wish to serve as prison guards, but that does not justify the exclusion of all women from this employment opportunity. Thus, “[i]n the usual case,” ibid., the Court’s interpretation of the bfoq exception would mandate hiring qualified women for guard jobs in maximum-security institutions. The highly successful experiences of other States allowing such job opportunities, see briefs for the States of California and Washington as amici curiae, confirm that absolute disqualification of women is not, in the words of Title VII, “reasonably necessary to the normal operation” of a maximum-security prison.

What would otherwise be considered unlawful discrimination against women is justified by the Court, however, on the *342basis of the “barbaric and inhumane” conditions in Alabama prisons, conditions so bad that state officials have conceded that they violate the Constitution. See Pugh v. Locke, 406 F. Supp. 318, 329, 331 (MD Ala. 1976). To me, this analysis sounds distressingly like saying two wrongs make a right. It is refuted by the plain words of § 703 (e). The statute requires that a bfoq be “reasonably necessary to the normal operation of that particular business or enterprise.” But no governmental “business” may operate “normally” in violation of the Constitution. Every action of government is constrained by constitutional limitations. While those limits may be violated more frequently than we would wish, no one disputes that the “normal operation” of all government functions takes place within them. A prison system operating in blatant violation of the Eighth Amendment is an exception that should be remedied with all possible speed, as Judge Johnson’s comprehensive order in Pugh v. Locke, supra, is designed to do. In the meantime, the existence of such violations should not be legitimatized by calling them “normal.” Nor should the Court accept them as justifying conduct that would otherwise violate a statute intended to remedy age-old discrimination.

The Court’s error in statutory construction is less objectionable, however, than the attitude it displays toward women. Though the Court recognizes that possible harm to women guards is an unacceptable reason for disqualifying women, it relies instead on an equally speculative threat to prison discipline supposedly generated by the sexuality of female guards. There is simply no evidence in the record to show that women guards would create any danger to security in Alabama prisons significantly greater than that which already exists. All of the dangers — -with one exception discussed below — are inherent in a prison setting, whatever the gender of the guards.

*343The Court first sees women guards as a threat to security because "there are few visible deterrents to inmate assaults on women custodians.” Ante, at 336. In fact, any prison guard is constantly subject to the threat of attack by inmates, and “invisible” deterrents are the guard's only real protection. No prison guard relies primarily on his or her ability to ward off an inmate attack to maintain order. Guards are typically unarmed and sheer numbers of inmates could overcome the normal complement. Rather, like all other law enforcement officers, prison guards must rely primarily on the moral authority of their office and the threat of future punishment for miscreants. As one expert testified below, common sense, fairness, and mental and emotional stability are the qualities a guard needs to cope with the dangers of the job. App. 81. Well qualified and properly trained women, no less than men, have these psychological weapons at their disposal.

The particular severity of discipline problems in the Alabama maximum-security prisons is also no justification for the discrimination sanctioned by the Court. The District Court found in Pugh v. Locke, supra, that guards “must spend all their time attempting to maintain control or to protect themselves.” 406 F. Supp., at 325. If male guards face an impossible situation, it is difficult to see how women could make the problem worse, unless one relies on precisely the type of generalized bias against women that the Court agrees Title VII was intended to outlaw. For example, much of the testimony of appellants’ witnesses ignores individual differences among members of each sex and reads like “ancient canards about the proper role of women.” Phillips v. Martin Marietta Corp., 400 U. S., at 545. The witnesses claimed that women guards are not strict disciplinarians; that they are physically less capable of protecting themselves and subduing unruly inmates; that inmates take advantage of them as they did their mothers, while male guards are strong father figures *344who easily maintain discipline, and so on.2 Yet the record shows that the presence of women guards has not led to a single incident amounting to a serious breach of security in any Alabama institution.3 And, in any event, “[g]uards rarely enter the cell blocks and dormitories,” Pugh v. Locke, 406 F. Supp., at 325, where the danger of inmate attacks is the greatest.

*345It appears that the real disqualifying factor in the Court’s view is “[t]he employee’s very womanhood.” Ante, at 336. The Court refers to the large number of sex offenders in Alabama prisons, and to “[t]he likelihood that inmates would assault a woman because she was a woman.” Ibid. In short, the fundamental justification for the decision is that women as guards will generate sexual assaults. With all respect, this rationale regrettably perpetuates one of the most insidious of the old myths about women — that women, wittingly or not, are seductive sexual objects. The effect of the decision, made I am sure with the best of intentions, is to punish women because their very presence might provoke sexual assaults. It is women who are made to pay the price in lost job opportunities for the threat of depraved conduct by prison inmates. Once again, “[t]he pedestal upon which women have been placed has . . . , upon closer inspection, been revealed as a cage.” Sail’er Inn, Inc. v. Kirby, 5 Cal. 3d 1, 20, 485 P. 2d 529, 541 (1971). It is particularly ironic that the cage is erected here in response to feared misbehavior by imprisoned criminals.4

The Court points to no evidence in the record to support the asserted “likelihood that inmates would assault a woman because she was a woman.” Ante, at 336. Perhaps the Court relies upon common sense, or “innate recognition,” Brief for Appellants 51. But the danger in this emotionally laden context is that common sense will be used to mask the “ ‘romantic paternalism’ ” and persisting discriminatory atti*346tudes that the Court properly eschews. Ante, at 335. To me, the only matter of innate recognition is that the incidence of sexually motivated attacks on guards will be minute compared to the “likelihood that inmates will assault” a guard because he or she is a guard.

The proper response to inevitable attacks on both female and male guards is not to limit the employment opportunities of law-abiding women who wish to contribute to their community, but to take swift and sure punitive action against the inmate offenders. Presumably, one of the goals of the Alabama prison system is the eradication of inmates’ antisocial behavior patterns so that prisoners will be able to live one day in free society. Sex offenders can begin this process by learning to relate to women guards in a socially acceptable manner. To deprive women of job opportunities because of the threatened behavior of convicted criminals is to turn our social priorities upside down.5

Although I do not countenance the sex discrimination *347condoned by the majority, it is fortunate that the Court’s decision is carefully limited to the facts before it. I trust the lower courts will recognize that the decision was impelled by the shockingly inhuman conditions in Alabama prisons, and thus that the “extremely narrow [bfoq] exception” recognized here, ante, at 334, will not be allowed “to swallow the rule” against sex discrimination. See Phillips v. Martin Marietta Corp., 400 U. S., at 545. Expansion of today’s decision beyond its narrow factual basis would erect a serious roadblock to economic equality for women.

concurring in No. 76-255 and dissent-

ing in No. 76-422.

I join the Court’s opinion in Hazelwood School Dist. v. United States, No. 76-255, ante, p. 299, but with reservations with respect to the relative neglect of applicant pool data in finding a prima facie case of employment discrimination and heavy reliance on the disparity between the areawide percentage of black public school teachers and the percentage of blacks on Hazelwood’s teaching staff. Since the issue is whether Hazelwood discriminated against blacks in hiring after Title VII became applicable to it in 1972, perhaps the Government should have looked initially to Hazelwood’s hiring practices in the 1972-1973 and 1973-1974 academic years with respect to the available applicant pool, rather than to history and to comparative work-force statistics from other school districts. Indeed, there is evidence in the record suggesting that Hazelwood, with a black enrollment of only 2%, hired a higher percentage of black applicants than of white applicants for these two years. The Court’s opinion, of course, permits Hazelwood to introduce applicant pool data on remand in order to rebut the prima facie case of a discriminatory pattern or practice. This may be the only fair and realistic allocation of the evidence burden, but arguably 'the United States should have been required to adduce evidence as to the applicant pool *348before it was entitled to its prima facie presumption. At least it might have been required to present some defensible ground for believing that the racial composition of Hazelwood’s applicant pool was roughly the same as that for the school districts in the general area, before relying on comparative work-force data to establish its prima facie case.

In Dothard v. Rawlinson, No. 76-422, I have more trouble agreeing that a prima facie case of sex discrimination was made out by statistics showing that the Alabama height and weight requirements would exclude a larger percentage of women in the United States than of men. As in Hazelwood, the issue is whether there was discrimination in dealing with actual or potential applicants; but in Hazelwood there was at least a colorable argument that the racial composition of the area-wide teacher work force was a reasonable proxy for the composition of the relevant applicant pool and hence that a large divergence between the percentage of blacks on the teaching staff and the percentage in the teacher work force raised a fair inference of racial discrimination in dealing with the applicant pool. In Dothard, however, I am unwilling to believe that the percentage of women applying or interested in applying for jobs as prison guards in Alabama approximates the percentage of women either in the national or state population. A plaintiff could, of course, show that the composition of the applicant pool was distorted by the exclusion of non-applicants who did not apply because of the allegedly discriminatory job requirement. But no such showing was made or even attempted here; and although I do not know what the actual fact is, I am not now convinced that a large percentage of the actual women applicants, or of those who are seriously interested in applying, for prison guard positions would fail to satisfy the height and weight requirements. Without a more satisfactory record on this issue, I cannot conclude that appel-lee Rawlinson has either made out a prima facie case for the invalidity of the restrictions or otherwise proved that she was *349improperly denied employment as a prison guard. There being no showing of discrimination, I do not reach the question of justification; nor, since she does not meet the threshold requirements for becoming a prison guard, need I deal with the gender-based requirements for contact positions. I dissent from the Court’s judgment in Dothard insofar as it affirms the judgment of the District Court.