delivered the opinion of the Court.

In Taylor v. Louisiana, 419 U. S. 522 (1975), this Court held that systematic exclusion of women during the jury-selection process, resulting in jury pools not “reasonably *359representative” of the community, denies a criminal defendant his right, under the Sixth and Fourteenth Amendments, to a petit jury selected from a fair cross section of the community.1 Under the system invalidated in Taylor, a woman could not serve on a jury unless she filed a written declaration of her willingness to do so.2 As a result, although 53% of the persons eligible for jury service were women, less than 1% of the 1,800 persons whose names were drawn from the jury wheel during the year in which appellant Taylor’s jury was chosen were female. Id., at 524.

At the time of our decision in Taylor, no other State provided that women could not serve on a jury unless they volunteered to serve.3 However, five States, including Missouri, provided an automatic exemption from jury service for any women requesting not to serve.4 Subsequent to Taylor, *360three of these States eliminated this exemption.5 Only Missouri, respondent in this case, and Tennessee6 continue to exempt women from jury service upon request.7 Today we hold that such systematic exclusion of women that results in jury venires averaging less than 15% female violates the Constitution’s fair-cross-section requirement.

I

Petitioner Duren was indicted in 1975 in the Circuit Court of Jackson County, Mo., for first-degree murder and first-degree robbery. In a pretrial motion to quash his petit jury panel, and again in a post-conviction motion for a new trial, he contended that his right to trial by a jury chosen from a fair cross section of his community was denied by provisions of Missouri law granting women who so request an automatic exemption from jury service.8 Both motions were denied.

*361At hearings on these motions, petitioner established that the jury-selection process in Jackson County begins with the annual mailing of a questionnaire to persons randomly selected from the Jackson County voter registration list. Approximately 70,000 questionnaires were mailed in 1975. The questionnaire contains a list of occupations and other categories which are the basis under Missouri law for either disqualification9 or exemption10 from jury service.11 Included on the questionnaire is a paragraph prominently addressed “TO WOMEN” that states in part:

“Any woman who elects not to serve will fill out this paragraph and mail this questionnaire to the jury commissioner at once.” 12

*362A similar paragraph is addressed "TO MEN OVER 65 YEARS OF AGE,” who are also statutorily exempt upon request.13

The names of those sent questionnaires are placed in the master jury wheel for Jackson County, except for those returning the questionnaire who indicate disqualification or claim an applicable exemption. Summonses are mailed on a weekly basis to prospective jurors randomly drawn from the jury wheel. The summons, like the questionnaire, contains special directions to men over 65 and to women, this time advising them to return the summons by mail if they desire not to serve. The practice also is that even those women who do not return the summons are treated as having claimed exemption if they fail to appear for jury service on the appointed day.14 Other persons seeking to claim an exemption at this stage must make written or personal application to the court.

Petitioner established that according to the 1970 census, 54% of the adult inhabitants of Jackson County were women. He also showed that for the periods June-October 1975 and January-March 1976,15 11,197 persons were summoned and that 2,992 of these, or 26.7%, were women. Of those summoned, 741 women and 4,378 men appeared for service. Thus, 14.5% (741 of 5,119) of the persons on the postsummons weekly venires during the period in which petitioner’s jury was chosen were female.16 In March 1976, when petitioner’s *363trial began, 15.5% of those on the weekly venires were women (110 of 707).17 Petitioner’s jury was selected from a 53-person panel on which there were 5 women; all 12 jurors chosen were men.18 None of the foregoing statistical evidence was disputed.

In affirming petitioner’s conviction, the Missouri Supreme Court questioned two aspects of his statistical presentation. First, it considered the census figures inadequate because they were six years old and might not precisely mirror the percentage of women registered to vote. Second, petitioner had not unequivocally demonstrated the extent to which the low percentage of women appearing for jury service was due to the automatic exemption for women, rather than to sex-neutral exemptions such as that for persons over age 65.

The court went on to hold, however, that even accepting petitioner’s statistical proof, “the number of female names in the wheel, those summoned and those appearing were well above acceptable constitutional standards.” 556 S. W. 2d 11, 15-17 (1977).19 We granted certiorari, 435 U. S. 1006 (1978), because of concern that the decision below is not consistent with our decision in Taylor.

II

We think that in certain crucial respects the Missouri Supreme Court misconceived the nature of the fair-cross-section inquiry set forth in Taylor. In holding that “petit juries must be drawn from a source fairly representative of the community,” 419 U. S., at 538, we explained that

“jury wheels, pools of names, panels, or venires from which juries are drawn must not systematically exclude *364distinctive groups in the community and thereby fail to be reasonably representative thereof.” Ibid.20

In order to establish a prima facie violation of the fair-cross-section requirement, the defendant must show (1) that the group alleged to be excluded is a “distinctive” group in the community; (2) that the representation of this group in venires from which juries are selected is not fair and reasonable in relation to the number of such persons in the community; and (3) that this underrepresentation is due to systematic exclusion of the group in the jury-selection process.

A

With respect to the first part of the prima facie test, Taylor without doubt established that women “are sufficiently numerous and distinct from men” so that “if they are systematically eliminated from jury panels, the Sixth Amendment’s fair-cross-section requirement. cannot be satisfied.” Id., at 531.

B

The second prong of the prima facie case was established by petitioner’s statistical presentation. Initially, the defendant must demonstrate the percentage of the community made up of the group alleged to be underrepresented, for this is the conceptual benchmark for the Sixth Amendment fair-cross-section requirement. In Taylor, the State had stipulated that 53% of the population eligible for jury service21 was female, while petitioner Duren has relied upon a census *365measurement of the actual percentage of women in the community (54%). In the trial court, the State of Missouri never challenged these data. Although the Missouri Supreme Court speculated that changing population patterns between 1970 and 1976 and unequal voter registration by men and women 22 rendered the census figures a questionable frame of reference,23 there is no evidence whatsoever in the record to suggest that the 1970 census data significantly distorted the percentage of women in Jackson County at the time of trial. Petitioner’s presentation was clearly adequate prima facie evidence of population characteristics for the purpose of making a fair-cross-section violation.24

Given petitioner’s proof that in the relevant community slightly over half of the adults are women, we must disagree with the conclusion of the court below that jury venires containing approximately 15% women are “reasonably rep*366resentative” of this community. If the percentage of women appearing on jury pools in Jackson County had precisely mirrored the percentage of women in the population, more than one of every two prospective jurors would have been female. In fact, less than one of every six prospective jurors was female; 85% of the average jury was male. Such a gross discrepancy between the percentage of women in jury venires and the percentage of women in the community requires the conclusion that women were not fairly represented in the source from which petit juries were drawn in Jackson County.

C

Finally, in order to establish a prima facie case, it was necessary for petitioner to show that the underrepresentation of women, generally and on his venire, was due to their systematic exclusion in the jury-selection process. Petitioner’s proof met this requirement. His undisputed demonstration that a large discrepancy occurred not just occasionally, but in every weekly venire for a period of nearly a year manifestly indicates that the cause of the underrepresentation was systematic — that is, inherent in the particular jury-selection process utilized.

Petitioner Duren’s statistics and other evidence also established when in the selection process the systematic exclusion took place. There was no indication that underrepresentation of women occurred at the first stage of the selection process— the questionnaire canvass of persons randomly selected from the relevant voter registration list. The first sign of a systematic discrepancy is at the next stage — the construction of the jury wheel from which persons are randomly summoned for service. Less than 30% of those summoned were female, demonstrating that a substantially larger number of women answering the questionnaire claimed either ineligibility or exemption from jury service. Moreover, at the summons stage women were not only given another opportunity to *367claim exemption, but also were presumed to have claimed exemption when they did not respond to the summons. Thus, the percentage of women at the final, venire, stage (14.5%) was much lower than the percentage of women who were summoned for service (26.7%).

The resulting disproportionate and consistent exclusion of women from the jury wheel and at the venire stage was quite obviously due to the system by which juries-were selected. Petitioner demonstrated that the underrepresentation of women in the final pool of prospective jurors was due to the operation of Missouri’s exemption criteria — whether the automatic exemption for women or other statutory exemptions — as implemented in Jackson County. Women were therefore systematically underrepresented within the meaning of Taylor.25

Ill

The demonstration of a prima facie fair-cross-section violar tion by the defendant is not the end of the inquiry into whether a constitutional violation has occurred. We have explained that “States remain free to prescribe relevant qualifications for their jurors and to provide reasonable exemptions so long as it may be fairly said that the jury lists or panels are representative of the community.” Taylor, 419 U. S., at 538. However, we cautioned that “[t]he right to a proper jury cannot be overcome on merely rational grounds,” id., at 534. Rather, it requires that a significant state interest be manifestly and primarily advanced by those aspects of the *368jury-selection process, such as exemption criteria, that result in the disproportionate exclusion of a distinctive group.26

The Supreme Court of Missouri suggested that the low-percentage of women on jury venires in Jackson County may have been due to a greater number of women than of men qualifying for or ‘claiming permissible exemptions, such as those for persons over 65, teachers, and government workers. 556 S. W. 2d, at 16. Respondent further argues that petitioner has not proved that the exemption for women had “any effect” on or was responsible for the underrepresentation of women on venires. Brief for Respondent 15.

However, once the defendant has made a prima facie showing of an infringement of his constitutional right to a jury drawn from a fair cross section of the community, it is the State that bears the burden of justifying this infringement by showing attainment of a fair cross section to be incompatible with a significant state interest. See Taylor, 419 U. S., at 533-535. Assuming, arguendo, that the exemptions mentioned *369by the court below would justify failure to achieve a fair community cross section on jury venires, the State must demonstrate that these exemptions caused the underrepresentation complained of. The record contains no such proof, and mere suggestions or assertions to that effect are insufficient.

The other possible cause of the disproportionate exclusion of women on Jackson County jury venires is, of course, the automatic exemption for women. Neither the Missouri Supreme Court nor respondent in its brief has offered any substantial justification for this exemption. In response to questioning at oral argument, counsel for respondent ventured that the only state interest advanced by the exemption is safeguarding the important role played by women in home and family fife.27 But exempting all women because of the preclusive domestic responsibilities of some women is insufficient justification for their disproportionate exclusion on jury venires. What we stated in Taylor with respect to the system there challenged under which women could “opt in” for jury service is equally applicable to Missouri’s “opt out” exemption:

“It is untenable to suggest these days that it would be a special hardship for each and every woman to perform jury service or that society cannot spare any women from their present duties. This may be the case with many, and it may be burdensome to sort out those who should be exempted from those who should serve. But that task is performed in the case of men, and the administrative convenience in dealing with women as a class is insufficient justification for diluting the quality of community judgment represented by the jury in criminal trials.

“If it was ever the case that women were unqualified to sit on juries or were so situated that none of them should be required to perform jury service, that time has long *370since passed.” 419 U. S., at 534-535, 537 (footnote omitted).

We recognize that a State may have an important interest in assuring that those members of the family responsible for the care of children are available to do so. An exemption appropriately tailored to this interest would, we think, survive a fair-cross-section challenge. We stress, however, that the constitutional guarantee to a jury drawn from a fair cross section of the community requires that States exercise proper caution in exempting broad categories of persons from jury service. Although most occupational and other reasonable exemptions may inevitably involve some degree of overinclu-siveness or underinclusiveness, any category expressly limited to a group in the community of sufficient magnitude and distinctiveness so as to be within the fair-cross-section requirement — such as women — runs the danger of resulting in under-representation sufficient to constitute a prima facie violation of that constitutional requirement. We also repeat the observation made in Taylor that it is unlikely that reasonable exemptions, such as those based on special hardship, incapacity, or community needs, “would pose substantial threats that the remaining pool of jurors would not be representative of the community.” Id., at 534.

The judgment of the Missouri Supreme Court is reversed, and the case is remanded for further proceedings not inconsistent with this opinion.

So ordered.

dissenting.

The Court steadfastly maintained in Taylor v. Louisiana, 419 U. S. 522 (1975), when it “distinguished” Hoyt v. Florida, 368 U. S. 57 (1961), that its'holding rested on the jury trial requirement of the Sixth and Fourteenth Amendments and not on the Equal Protection Clause of the Fourteenth Amendment. Today’s decision makes a halfhearted effort to con*371tinue that fiction in footnotes 1 and 26, declaring that cases based on the Equal Protection Clause, such as Alexander v. Louisiana, 405 U. S. 625 (1972), are not “entirely analogous” to the case at hand. The difference apparently lies in the fact, among others, that under equal protection analysis prima facie challenges are rebuttable by proof of absence of intent to discriminate, while under Sixth Amendment analysis intent is irrelevant, but the State may show “adequate justification” for the disproportionate representation of the classes being compared. We are reminded, however, that disproportion-ality may not be justified “on merely rational grounds” and that justification requires that “a significant state interest be manifestly and primarily advanced” by the exemption criteria resulting in the disproportionate representation. Ante, at 367 (emphasis supplied). That this language has strong overtones of equal protection is demonstrated in this Court’s most recent application of the Equal Protection Clause to distinctions between men and women: “ ‘[Classifications by gender must serve important governmental objectives and must be substantially related to the achievement of those objectives.’ ” Califano v. Goldfarb, 430 U. S. 199, 210-211 (1977) (plurality opinion), quoting Craig v. Boren, 429 U. S. 190, 197 (1976) (emphasis supplied). The Constitution does not require, and our jurisprudence is ill served, by a hybrid doctrine such as that developed in Taylor, and in this ease.*

*372Even if I were able to reconcile the Court’s agile amalgamation of the Due Process Clause and the Equal Protection Clause of the Fourteenth Amendment in deciding this case *373and Taylor, I have no little concern about where the road upon which the Court has embarked will ultimately lead. In Taylor, the Court relied upon cases dealing with outright exclusion of racial groups, Smith v. Texas, 311 U. S. 128 (1940), and of women, Ballard v. United States, 329 U. S. 187 (1946), from jury service. Although in Smith, the exclusion had been covert, in Ballard the exclusion had been overt. The Court in Taylor concluded, I assume on the basis of these cases, that “women cannot be systematically excluded from jury panels from which petit juries are drawn.” 419 U. S., at 533.

In Taylor, as in Hoyt v. Florida, 368 U. S. 57 (1961), *374women had not been actually prohibited or excluded from serving on juries. But requirements, inapplicable to men, that they affirmatively make known to the jury commissioner their desire to serve had for all practical purposes had that effect. Indeed, in Taylor not one woman appeared on a venire of 175 persons drawn for jury service in the parish in question. 419 U. S., at 524. Taylor, by its language and on its facts, was an “exclusion” case.

Here, on the other hand, the Court in one sentence both asserts that it can, and admits that it cannot, treat the system used in Jackson County, Mo., as one which “excludes” women, saying: “Today we hold that such systematic exclusion of women that results in jury venires averaging less than 15% female violates the Constitution’s fair-cross-section requirement.” Ante, at 360. If there are indeed 15% women on the jury panels in Jackson County, the Court uses the word “exclusion” contrary to any use of the word with which I am familiar. Women are undoubtedly underrepresented as compared to men on Jackson County juries, but therein lies the difference between this case and Taylor.

Eventually the Court either will insist that women be treated identically to men for purposes of jury selection (which is intimated in dicta, ante, at 365-366, 370), or in some later sequel to this line of cases will discover some peculiar magic in the number 15 that will enable it to distinguish between such a percentage and higher percentages less than 50. But whichever of these routes the Court chooses to travel when the question is actually presented, its decision today puts state legislators and local jury commissioners at a serious disadvantage wholly unwarranted by the constitutional provisions upon which it relies. If the Court ultimately concludes that men and women must be treated exactly alike for purposes of jury service, it will have imposed substantial burdens upon many women, particularly in less populated areas, without necessarily producing any corresponding increase in the repre*375sentative character of jury panels. If it ultimately concludes that a percentage of women on jury panels greater than 15 but substantially less than 50 is permissible even though the State’s jury selection system permits women but not men to “opt out” of jury service, it is simply playing a constitutional numbers game.

The attorneys general and prosecuting attorneys in the various States, sensibly concluding that a 15% representation of women on jury venires cannot in any rational legal system be materially different from a 20% representation, will press legislators and jury commissioners to abolish all distinctions between men and women for purposes of jury service. Understandably unhappy with the prospect of having still more convictions for armed robbery or murder set aside at the behest of male defendants claiming that women were insufficiently represented on their jury panel, these state attorneys will make their informed but inevitably parochial views known in the halls of their respective legislatures. These views will presumably be in harmony with those of the organized women’s groups that have appeared as amici curiae 'in similar cases, asserting that the Constitution prohibits women from being given a choice as to whether they will serve on juries when men are required to serve.

Nor are distinctions between men and women in jury selections likely to be the only casualties to result from today’s opinion. Apparently realizing the desirability of some predictability if otherwise fairly tried defendants are to be freed on the basis of such a constitutional numbers game, the Court ventures the view that an “exemption appropriately tailored” to the State’s interest in ensuring that those members of the family responsible for the care of children are available to perform such care would “survive a fair-cross-section challenge.” Ante, at 370. It also repeats the “observation” made in Taylor that it is “unlikely that reasonable exemptions, such as those based on special hardship, inca*376pacity, or community needs, 'would pose substantial threats that the remaining pool of jurors would not be representative of the community.’ ” Ibid. But the States are warned that the Constitution requires them to “exercise proper caution in exempting broad categories of persons from jury service,” even though “mos^ occupational and other reasonable exemptions may inevitably involve some degree of overinclusiveness or underinclusiveness . . . Ibid.

The lot of a legislator or judge attempting to conform a State’s jury selection process to the dictates of today’s opinion, and yet recognize what may be very valid state interests in excusing some individuals or classes of individuals from jury service, is surely not a happy one. Will the Court’s above-quoted dicta soon meet the same fate that the decision in Hoyt v. Florida, supra, met in Taylor, or will they survive longer?

There is more than adequate documentation for the proposition that jury service is not a pleasant experience in many jurisdictions and that it tends to be time consuming and often seemingly useless from the point of view of the prospective juror. To the extent that States may engage in the process of jury selection by broad classifications, and by a system of exemptions which require a minimum of administrative effort, the frustrations of jury service will be at least in part alleviated, and perhaps the Court’s stated goal of a “fair cross section” actually advanced. On the other hand, to the extent that such forms of selection are deemed constitutionally impermissible, and case-by-case “opting out” required with respect to each prospective juror, the ordeal of the prospective juror becomes more burdensome, and the State’s administrative task more time consuming. Since most States will undoubtedly wish to immunize otherwise valid criminal convictions against reversal on the basis of the Court’s most recent exegesis of the Fourteenth Amendment’s requirements on the jury selection process, their natural tendency will be *377to impose these burdens on citizen jurors and judicial administrators in order to avoid any possibility of a successful constitutional attack on the composition of the jury.

The probability, then, is that today’s decision will cause States to abandon not only gender-based but also occupation-based classifications for purposes of jury service. Doctors and nurses, though virtually irreplaceable in smaller communities, may ultimately be held by the Court to bring their own “flavor” or “indefinable something” to a jury venire. See supra, at 372 n. If so, they could then be exempted from jury service only on a case-by-case basis, and would join others with skills much less in demand whiling away their time in jury rooms of countless courthouses.

No one but a lawyer could think that this was a managerially sound solution to an important problem of judicial administration, and no one but a lawyer thoroughly steeped in the teachings of cases such as Taylor, Goldfarb, and Craig could think that such a solution was mandated by the United States Constitution. No large group of people can be conscripted to serve on juries nationwide, any more than in armies, without the use of broad general classifications which may not fit in every case the purpose for which the classification was designed. The alternative is case-by-case treatment which entails administrative burdens out of all proportion to the end sought to be achieved.

The short of it is that the only winners in today’s decision are those in the category of petitioner, now freed of his conviction of first-degree murder. They are freed not because of any demonstrable unfairness at any stage of their trials, but because of the Court’s obsession that criminal venires represent a “fair cross section” of the community, whatever that may be. The losers are the remaining members of that community — men and women seeking to do their duty as jurors and yet minimize the inconvenience that such service entails, judicial administrators striving to make *378the criminal justice system function, and the citizenry in general seeking the incarceration of those convicted of serious crimes after a fair trial. I do not believe that the Fourteenth Amendment was intended or should be interpreted to produce such a quixotic result.