delivered the opinion of the Court.
This case requires us to determine the constitutionality of regulations promulgated by the Missouri Division of Corrections relating to inmate marriages and inma.te-to-inma.te correspondence. The Court of Appeals for the Eighth Circuit, applying a strict scrutiny analysis, concluded that the regulations violate respondents’ constitutional rights. We hold that a lesser standard of scrutiny is appropriate in determining the constitutionality of the prison rules. Applying that standard, we uphold the validity of the correspondence regulation, but we conclude that the marriage restriction cannot be sustained.
H
Respondents brought this class action for injunctive relief and damages in the United States District Court for the Western District of Missouri. The regulations challenged in the complaint were in effect at all prisons within the jurisdiction of the Missouri Division of Corrections. This litigation focused, however, on practices at the Renz Correctional Institution (Renz), located in Cedar City, Missouri. The Renz prison population includes both male and female prisoners of varying security levels. Most of the female prisoners at Renz are classified as medium or maximum security inmates, while most of the male prisoners are classified as minimum security offenders. Renz is used on occasion to provide protective custody for inmates from other prisons in the Missouri system. The facility originally was built as a minimum security prison farm, and it still has a minimum security perimeter without guard towers or walls.
Two regulations are at issue here. The first of the challenged regulations relates to correspondence between inmates at different institutions. It permits such correspondence “with immediate family members who are inmates in other correctional institutions,” and it permits correspondence between inmates “concerning legal matters.” Other correspondence between inmates, however, is permitted only *82if “the classification/treatment team of each inmate deems it in the best interest of the parties involved.” App. 34. Trial testimony indicated that as a matter of practice, the determination whether to permit inmates to correspond was based on team members’ familiarity with the progress reports, conduct violations, and psychological reports in the inmates’ files rather than on individual review of each piece of mail. See 777 F. 2d 1307, 1308 (CA8 1985). At Renz, the District Court found that the rule “as practiced is that inmates may not write non-family inmates.” 586 F. Supp. 589, 591 (WD Mo. 1984).
The challenged marriage regulation, which was promulgated while this litigation was pending, permits an inmate to marry only with the permission of the superintendent of the prison, and provides that such approval should be given only “when there are compelling reasons to do so.” App. 47. The term “compelling” is not defined, but prison officials testified at trial that generally only a pregnancy or the birth of an illegitimate child would be considered a compelling reason. See 586 F. Supp., at 592. Prior to the promulgation of this rule, the applicable regulation did not obligate Missouri Division of Corrections officials to assist an inmate who wanted to get married, but it also did not specifically authorize the superintendent of an institution to prohibit inmates from getting married. Ibid.
The District Court certified respondents as a class pursuant to Federal Rule of Civil Procedure 23. The class certified by the District Court includes “persons who either are or may be confined to the Renz Correctional Center and who desire to correspond with inmates at other Missouri correctional facilities.” It also encompasses a broader group of persons “who desire to . . . marry inmates of Missouri correctional institutions and whose rights of . . . marriage have been or will be violated by employees of the Missouri Division of Corrections.” See App. 21-22.
*83The District Court issued a memorandum opinion and order finding both the correspondence and marriage regulations unconstitutional. The court, relying on Procunier v. Martinez, 416 U. S. 396, 413-414 (1974), applied a strict scrutiny standard. It held the marriage regulation to be an unconstitutional infringement upon the fundamental right to marry because it was far more restrictive than was either reasonable or essential for the protection of the State’s interests in security and rehabilitation. 586 F. Supp., at 594. The correspondence regulation also was unnecessarily broad, the court concluded, because prison officials could effectively cope with the security problems raised by inmate-to-inmate correspondence through less restrictive means, such as scanning the mail of potentially troublesome inmates. Id., at 596. The District Court also held that the correspondence regulation had been applied in an arbitrary and capricious manner.
The Court of Appeals for the Eighth Circuit affirmed. 777 F. 2d 1307 (1985). The Court of Appeals held that the District Court properly used strict scrutiny in evaluating the constitutionality of the Missouri correspondence and marriage regulations. Under Procunier v. Martinez, supra, the correspondence regulation could be justified “only if it furthers an important or substantial governmental interest unrelated to the suppression of expression, and the limitation is no greater than necessary or essential to protect that interest.” 777 F. 2d, at 1310. The correspondence regulation did not satisfy this standard because it was not the least restrictive means of achieving the security goals of the regulation. In the Court of Appeals’ view, prison officials could meet the problem of inmate conspiracies by exercising their authority to open and read all prisoner mail. Id., at 1315-1316. The Court of Appeals also concluded that the marriage rule was not the least restrictive means of achieving the asserted goals of rehabilitation and security. The goal of rehabilitation could be met through alternatives such *84as counseling, and violent “love triangles” were as likely to occur without a formal marriage ceremony as with one. Ibid. Absent evidence that the relationship was or would become abusive, the connection between an inmate’s marriage and the subsequent commission of a crime was simply too tenuous to justify denial of this constitutional right. Id., at 1315.
We granted certiorari, 476 U. S. 1139 (1986).
HH HH
We begin, as did the courts below, with our decision in Procunier v. Martinez, supra, which described the principles that necessarily frame our analysis of prisoners’ constitutional claims. The first of these principles is that federal courts must take cognizance of the valid constitutional claims of prison inmates. Id., at 405. Prison walls do not form a barrier separating prison inmates from the protections of the Constitution. Hence, for example, prisoners retain the constitutional right to petition the government for the redress of grievances, Johnson v. Avery, 393 U. S. 483 (1969); they are protected against invidious racial discrimination by the Equal Protection Clause of the Fourteenth Amendment, Lee v. Washington, 390 U. S. 333 (1968); and they enjoy the protections of due process, Wolff v. McDonnell, 418 U. S. 539 (1974); Haines v. Kerner, 404 U. S. 519 (1972). Because prisoners retain these rights, “[w]hen a prison regulation or practice offends a fundamental constitutional guarantee, federal courts will discharge their duty to protect constitutional rights.” Procunier v. Martinez, 416 U. S., at 405-406.
A second principle identified in Martinez, however, is the recognition that “courts are ill equipped to deal with the increasingly urgent problems of prison administration and reform.” Id., at 405. As the Martinez Court acknowledged, “the problems of prisons in America are complex and intractable, and, more to the point, they are not readily susceptible of resolution by decree. ” Id., at 404-405. Running a prison *85is an inordinately difficult undertaking that requires expertise, planning, and the commitment of resources, all of which are peculiarly within the province of the legislative and executive branches of government. Prison administration is, moreover, a task that has been committed to the responsibility of those branches, and separation of powers concerns counsel a policy of judicial restraint. Where a state penal system is involved, federal courts have, as we indicated in Martinez, additional reason to accord deference to the appropriate prison authorities. See id., at 405.
Our task, then, as we stated in Martinez, is to formulate a standard of review for prisoners’ constitutional claims that is responsive both to the “policy of judicial restraint regarding prisoner complaints and [to] the need to protect constitutional rights.” Id., at 406. As the Court of Appeals acknowledged, Martinez did not itself resolve the question that it framed. Martinez involved mail censorship regulations proscribing statements that “unduly complain,” “magnify grievances,” or express “inflammatory political, racial, religious or other views.” Id., at 415. In that case, the Court determined that the proper standard of review for prison restrictions on correspondence between prisoners and members of the general public could be decided without resolving the “broad questions of ‘prisoners’ rights.’” Id., at 408. The Martinez Court based its ruling striking down the content-based regulation on the First Amendment rights of those who are not prisoners, stating that “[wjhatever the status of a prisoner’s claim to uncensored correspondence with an outsider, it is plain that the latter’s interest is grounded in the First Amendment’s guarantee of freedom of speech.” Id., at 408. Our holding therefore turned on the fact that the challenged regulation caused a “consequential restriction on the First and Fourteenth Amendment rights of those who are not prisoners.” Id., at 409 (emphasis added). We expressly reserved the question of the proper standard of *86review to apply in cases “involving questions of ‘prisoners’ rights.’” Ibid.
In four cases following Martinez, this Court has addressed such “questions of ‘prisoners’ rights.’” The first of these, Pell v. Procunier, 417 U. S. 817 (1974), decided the same Term as Martinez, involved a constitutional challenge to a prison regulation prohibiting face-to-face media interviews with individual inmates. The Court rejected the inmates’ First Amendment challenge to the ban on media interviews, noting that judgments regarding prison security “are peculiarly within the province and professional expertise of corrections officials, and, in the absence of substantial evidence in the record to indicate that the officials have exaggerated their response to these considerations, courts should ordinarily defer to their expert judgment in such matters.” 417 U. S., at 827.
The next case to consider a claim of prisoners’ rights was Jones v. North Carolina Prisoners’ Union, 433 U. S. 119 (1977). There the Court considered prison regulations that prohibited meetings of a “prisoners’ labor union,” inmate solicitation of other inmates to join the union, and bulk mailings concerning the union from outside sources. Noting that the lower court in Jones had “got[ten] off on the wrong foot. . . by not giving appropriate deference to the decisions of prison administrators and appropriate recognition to the peculiar and restrictive circumstances of penal confinement,” id., at 125, the Court determined that the First and Fourteenth Amendment rights of prisoners were “barely implicated” by the prohibition on bulk mailings, see id., at 130, and that the regulation was “reasonable” under the circumstances. The prisoners’ constitutional challenge to the union meeting and solicitation restrictions was also rejected, because “[t]he ban on inmate solicitation and group meetings . . . was rationally related to the reasonable, indeed to the central, objectives of prison administration.” Id., at 129.
*87Bell v. Wolfish, 441 U. S. 520 (1979), concerned a First Amendment challenge to a Bureau of Prisons rule restricting inmates’ receipt of hardback books unless mailed directly from publishers, book clubs, or bookstores. The rule was upheld as a “rational response” to a clear security problem. Id., at 550. Because there was “no evidence” that officials had exaggerated their response to the security problem, the Court held that “the considered judgment of these experts must control in the absence of prohibitions far more sweeping than those involved here.” Id., at 551. And in Block v. Rutherford, 468 U. S. 576 (1984), a ban on contact visits was upheld on the ground that “responsible, experienced administrators have determined, in their sound discretion, that such visits will jeopardize the security of the facility,” and the regulation was “reasonably related” to these security concerns. Id., at 589, 586.
In none of these four “prisoners’ rights” cases did the Court apply a standard of heightened scrutiny, but instead inquired whether a prison regulation that burdens fundamental rights is “reasonably related” to legitimate penological objectives, or whether it represents an “exaggerated response” to those concerns. The Court of Appeals in this case nevertheless concluded that Martinez provided the closest analogy for determining the appropriate standard of review for resolving respondents’ constitutional complaints. The Court of Appeals distinguished this Court’s decisions in Pell, Jones, Bell, and Block as variously involving “time, place, or manner” regulations, or regulations that restrict “presumptively dangerous” inmate activities. See 777 F. 2d, at 1310-1312. The Court of Appeals acknowledged that Martinez had expressly reserved the question of the appropriate standard of review based on inmates’ constitutional claims, but it nonetheless believed that the Martinez standard was the proper one to apply to respondents’ constitutional claims.
We disagree with the Court of Appeals that the reasoning in our cases subsequent to Martinez can be so narrowly *88cabined. In Pell, for example, it was found “relevant” to the reasonableness of a restriction on face-to-face visits between prisoners and news reporters that prisoners had other means of communicating with members of the general public. See 417 U. S., at 823-824. These alternative means of communication did not, however, make the prison regulation a “time, place, or manner” restriction in any ordinary sense of the term. As Pell acknowledged, the alternative methods of personal communication still available to prisoners would have been “unimpressive” if offered to justify a restriction on personal communication among members of the general public. Id., at 825. Nevertheless, they were relevant in determining the scope of the burden placed by the regulation on inmates’ First Amendment rights. Pell thus simply teaches that it is appropriate to consider the extent of this burden when “we [are] called upon to balance First Amendment rights against [legitimate] governmental interests.” Id., at 824.
Nor, in our view, can the reasonableness standard adopted in Jones and Bell be construed as applying only to “presumptively dangerous” inmate activities. To begin with, the Court of Appeals did not indicate how it would identify such “presumptively dangerous” conduct, other than to conclude that the group meetings in Jones, and the receipt of hardback books in Bell, both fall into that category. See 777 F. 2d, at 1311-1312. The Court of Appeals found that correspondence between inmates did not come within this grouping because the court did “not think a letter presents the same sort of ‘obvious security problem’ as does a hardback book.” Id., at 1312. It is not readily apparent, however, why hardback books, which can be scanned for contraband by electronic devices and fluoroscopes, see Bell v. Wolfish, supra, at 574 (Marshall, J., dissenting), are qualitatively different in this respect from inmate correspondence, which can be written in codes not readily subject to detection; or why coordinated inmate activity within the same prison is categorically different *89from inmate activity coordinated by mail among different prison institutions. The determination that an activity is “presumptively dangerous” appears simply to be a conclusion about the reasonableness of the prison restriction in light of the articulated security concerns. It therefore provides a tenuous basis for creating a hierarchy of standards of review.
If Pell, Jones, and Bell have not already resolved the question posed in Martinez, we resolve it now: when a prison regulation impinges on inmates’ constitutional rights, the regulation is valid if it is reasonably related to legitimate peno-logical interests. In our view, such a standard is necessary if “prison administrators . . . , and not the courts, [are] to make the difficult judgments concerning institutional operations.” Jones v. North Carolina Prisoners’ Union, 433 U. S., at 128. Subjecting the day-to-day judgments of prison officials to an inflexible strict scrutiny analysis would seriously hamper their ability to anticipate security problems and to adopt innovative solutions to the intractable problems of prison administration. The rule would also distort the decisionmaking process, for every administrative judgment would be subject to the possibility that some court somewhere would conclude that it had a less restrictive way of solving the problem at hand. Courts inevitably would become the primary arbiters of what constitutes the best solution to every administrative problem, thereby “unnecessarily perpetuating] the involvement of the federal courts in affairs of prison administration.” Procunier v. Martinez, 416 U. S., at 407.
As our opinions in Pell, Bell, and Jones show, several factors are relevant in determining the reasonableness of the regulation at issue. First, there must be a “valid, rational connection” between the prison regulation and the legitimate governmental interest put forward to justify it. Block v. Rutherford, supra, at 586. Thus, a regulation cannot be sustained where the logical connection between the regulation and the asserted goal is so remote as to render the policy *90arbitrary or irrational. Moreover, the governmental objective must be a legitimate and neutral one. We have found it important to inquire whether prison regulations restricting inmates’ First Amendment rights operated in a neutral fashion, without regard to the content of the expression. See Pell v. Procunier, 417 U. S., at 828; Bell v. Wolfish, 441 U. S., at 551.
A second factor relevant in determining the reasonableness of a prison restriction, as Pell shows, is whether there are alternative means of exercising the right that remain open to prison inmates. Where “other avenues” remain available for the exercise of the asserted right, see Jones v. North Carolina Prisoners’ Union, supra, at 131, courts should be particularly conscious of the “measure of judicial deference owed to corrections officials ... in gauging the validity of the regulation.” Pell v. Procunier, supra, at 827.
A third consideration is the impact accommodation of the asserted constitutional right will have on guards and other inmates, and on the allocation of prison resources generally. In the necessarily closed environment of the correctional institution, few changes will have no ramifications on the liberty of others or on the use of the prison’s limited resources for preserving institutional order. When accommodation of an asserted right will have a significant “ripple effect” on fellow inmates or on prison staff, courts should be particularly deferential to the informed discretion of corrections officials. Cf. Jones v. North Carolina Prisoners’ Union, supra, at 132-133.
Finally, the absence of ready alternatives is evidence of the reasonableness of a prison regulation. See Block v. Rutherford, 468 U. S., at 587. By the same token, the existence of obvious, easy alternatives may be evidence that the regulation is not reasonable, but is an “exaggerated response” to prison concerns. This is not a “least restrictive alternative” test: prison officials do not have to set up and then shoot down every conceivable alternative method of accommodat*91ing the claimant’s constitutional complaint. See ibid. But if an inmate claimant can point to an alternative that fully accommodates the prisoner’s rights at de minimis cost to valid penological interests, a court may consider that as evidence that the regulation does not satisfy the reasonable relationship standard.
Ill
Applying our analysis to the Missouri rule barring inmate-to-inmate correspondence, we conclude that the record clearly demonstrates that the regulation was reasonably related to legitimate security interests. We find that the marriage restriction, however, does not satisfy the reasonable relationship standard, but rather constitutes an exaggerated response to petitioners’ rehabilitation and security concerns.
A
According to the testimony at trial, the Missouri correspondence provision was promulgated primarily for security reasons. Prison officials testified that mail between institutions can be used to communicate escape plans and to arrange assaults and other violent acts. 2 Tr. 76; 4 id., at 225-228. Witnesses stated that the Missouri Division of Corrections had a growing problem with prison gangs, and that restricting communications among gang.members, both by transferring gang members to different institutions and by restricting their correspondence, was an important element in combating this problem. 2 id., at 75-77; 3 id., at 266-267; 4 id., at 226. Officials also testified that the use of Renz as a facility to provide protective custody for certain inmates could be compromised by permitting correspondence between inmates at Renz and. inmates at other correctional institutions. 3 id., at 264-265.
The prohibition on correspondence between institutions is logically connected to these legitimate security concerns. Undoubtedly, communication with other felons is a potential spur to criminal behavior: this sort of contact frequently is *92prohibited even after an inmate has been released on parole. See, e. g., 28 CFR §2.40(a)(10) (1986) (federal parole conditioned on nonassociation with known criminals, unless permission is granted by the parole officer). In Missouri prisons, the danger of such coordinated criminal activity is exacerbated by the presence of prison gangs. The Missouri policy of separating and isolating gang members — a strategy that has been frequently used to control gang activity, see G. Camp & C. Camp, U. S. Dept, of Justice, Prison Gangs: Their Extent, Nature and Impact on Prisons 64-65 (1985)— logically is furthered by the restriction on prisoner-to-prisoner correspondence. Moreover, the correspondence regulation does not deprive prisoners of all means of expression. Rather, it bars communication only with a limited class of other people with whom prison officials have particular cause to be concerned — inmates at other institutions within the Missouri prison system.
We also think that the Court of Appeals’ analysis overlooks the impact of respondents’ asserted right on other inmates and prison personnel. Prison officials have stated that in their expert opinion, correspondence between prison institutions facilitates the development of informal organizations that threaten the core functions of prison administration, maintaining safety and internal security. As a result, the correspondence rights asserted by respondents, like the organizational activities at issue in Jones v. North Carolina Prisoners’ Union, 433 U. S. 119 (1977), can be exercised only at the cost of significantly less liberty and safety for everyone else, guards and other prisoners alike. Indeed, the potential “ripple effect” is even broader here than in Jones, because exercise of the right affects the inmates and staff of more than one institution. Where exercise of a right requires this kind of tradeoff, we think that the choice made by corrections officials — which is, after all, a judgment “peculiarly within [their] province and professional expertise,” Pell v. Pro*93cunier, 417 U. S., at 827—should not be lightly set aside by the courts.
Finally, there are no obvious, easy alternatives to the policy adopted by petitioners. Other well-run prison systems, including the Federal Bureau of Prisons, have concluded that substantially similar restrictions on inmate correspondence were necessary to protect institutional order and security. See, e. g., 28 CFR §540.17 (1986). As petitioners have shown, the only alternative proffered by the claimant prisoners, the monitoring of inmate correspondence, clearly would impose more than a de minimis cost on the pursuit of legitimate corrections goals. Prison officials testified that it would be impossible to read every piece of inmate-to-inmate correspondence, 8 Tr. 159, 4 id., at 42-43, and consequently there would be an appreciable risk of missing dangerous messages. In any event, prisoners could easily write in jargon or codes to prevent detection of their real messages. See Camp & Camp, supra, at 130 (noting “frequent” use of coded correspondence by gang members in federal prison); see also Brief for State of Texas as Amicus Curiae 7-9. The risk of missing dangerous communications, taken together with the sheer burden on staff resources required to conduct item-by-item censorship, see 3 Tr. 176, supports the judgment of prison officials that this alternative is not an adequate alternative to restricting correspondence.
The prohibition on correspondence is reasonably related to valid corrections goals. The rule is content neutral, it logically advances the goals of institutional security and safety identified by Missouri prison officials, and it is not an exaggerated response to those objectives. On that basis, we conclude that the regulation does not unconstitutionally abridge the First Amendment rights of prison inmates.*
*94B
In support of the marriage regulation, petitioners first suggest that the rule does not deprive prisoners of a constitu*95tionally protected right. They concede that the decision to marry is a fundamental right under Zablocki v. Redhail, 434 U. S. 374 (1978), and Loving v. Virginia, 388 U. S. 1 (1967), but they imply that a different rule should obtain “in ... a prison forum.” See Brief for Petitioners 38, n. 6. Petitioners then argue that even if the regulation burdens inmates’ constitutional rights, the restriction should be tested under a reasonableness standard. They urge that the restriction is reasonably related to legitimate security and rehabilitation concerns.
We disagree with petitioners that Zablocki does not apply to prison inmates. It is settled that a prison inmate “retains those [constitutional] rights that are not inconsistent with his status as a prisoner or with the legitimate penological objectives of the corrections system.” Pell v. Procunier, supra, at 822. The right to marry, like many other rights, is subject to substantial restrictions as a result of incarceration. Many important attributes of marriage remain, however, after taking into account the limitations imposed by prison life. First, inmate marriages, like others, are expressions of emotional support and public commitment. These elements *96are an important and significant aspect of the marital relationship. In addition, many religions recognize marriage as having spiritual significance; for some inmates and their spouses, therefore, the commitment of marriage may be an exercise of religious faith as well as an expression of personal dedication. Third, most inmates eventually will be released by parole or commutation, and therefore most inmate marriages are formed in the expectation that they ultimately will be fully consummated. Finally, marital status often is a precondition to the receipt of government benefits [e. g., Social Security benefits), property rights (e. g., tenancy by the entirety, inheritance rights), and other, less tangible benefits (e. g., legitimation of children born out of wedlock). These incidents of marriage, like the religious and personal aspects of the marriage commitment, are unaffected by the fact of confinement or the pursuit of legitimate corrections goals.
Taken together, we conclude that these remaining elements are sufficient to form a constitutionally protected marital relationship in the prison context. Our decision in Butler v. Wilson, 415 U. S. 953 (1974), summarily affirming Johnson v. Rockefeller, 365 F. Supp. 377 (SDNY 1973), is not to the contrary. That case involved a prohibition on marriage only for inmates sentenced to life imprisonment; and, importantly, denial of the right was part of the punishment for crime. See id., at 381-382 (Lasker, J., concurring in part and dissenting in part) (asserted governmental interest of punishing crime sufficiently important to justify deprivation of right); see generally Mandel v. Bradley, 432 U. S. 173, 176 (1977) (“Because a summary affirmance is an affirmance of the judgment only, the rationale of the affirmance may not be gleaned solely from the opinion below”).
The Missouri marriage regulation prohibits inmates from marrying unless the prison superintendent has approved the marriage after finding that there are compelling reasons for doing so. As noted previously, generally only pregnancy or birth of a child is considered a “compelling reason” to approve *97a marriage. In determining whether this regulation imper-missibly burdens the right to marry, we note initially that the regulation prohibits marriages between inmates and civilians, as well as marriages between inmates. See Brief for Petitioners 40. Although not urged by respondents, this implication of the interests of nonprisoners may support application of the Martinez standard, because the regulation may entail a “consequential restriction on the [constitutional] rights of those who are not prisoners.” See Procunier v. Martinez, 416 U. S., at 409. We need not reach this question, however, because even under the reasonable relationship test, the marriage regulation does not withstand scrutiny.
Petitioners have identified both security and rehabilitation concerns in support of the marriage prohibition. The security concern emphasized by petitioners is that “love triangles” might lead to violent confrontations between inmates. See Brief for Petitioners 13, 36, 39. With respect to rehabilitation, prison officials testified that female prisoners often were subject to abuse at home or were overly dependent on male figures, and that this dependence or abuse was connected to the crimes they had committed. 3 Tr. 164-165. The superintendent at Renz, petitioner William Turner, testified that in his view, these women prisoners needed to concentrate on developing skills of self-reliance, 1 id., at 80-81, and that the prohibition on marriage furthered this rehabilitative goal. Petitioners emphasize that the prohibition on marriage should be understood in light of Superintendent Turner’s experience with several ill-advised marriage requests from female inmates. Brief for Petitioners 32-34.
We conclude that on this record, the Missouri prison regulation, as written, is not reasonably related to these penological interests. No doubt legitimate security concerns may require placing reasonable restrictions upon an inmate’s right to marry, and may justify requiring approval of the superintendent. The Missouri regulation, however, represents an *98exaggerated response to such security objectives. There are obvious, easy alternatives to the Missouri regulation that accommodate the right to marry while imposing a de minimis burden on the pursuit of security objectives. See, e. g., 28 CFR §551.10 (1986) (marriage by inmates in federal prison generally permitted, but not if warden finds that it presents a threat to security or order of institution, or to public safety). We are aware of no place in the record where prison officials testified that such ready alternatives would not fully satisfy their security concerns. Moreover, with respect to the security concern emphasized in petitioners’ brief — the creation of “love triangles” — petitioners have pointed to nothing in the record suggesting that the marriage regulation was viewed as preventing such entanglements. Common sense likewise suggests that there is no logical connection between the marriage restriction and the formation of love triangles: surely in prisons housing both male and female prisoners, inmate rivalries are as likely to develop without a formal marriage ceremony as with one. Finally, this is not an instance where the “ripple effect” on the security of fellow inmates and prison staff justifies a broad restriction on inmates’ rights — indeed, where the inmate wishes to marry a civilian, the decision to marry (apart from the logistics of the wedding ceremony) is a completely private one.
Nor, on this record, is the marriage restriction reasonably related to the articulated rehabilitation goal. First, in requiring refusal of permission absent a finding of a compelling reason to allow the marriage, the rule sweeps much more broadly than can be explained by petitioners’ penological objectives. Missouri prison officials testified that generally they had experienced no problem with the marriage of male inmates, see, e. g., 2 Tr. 21-22, and the District Court found that such marriages had routinely been allowed as a matter of practice at Missouri correctional institutions prior to adoption of the rule, 586 F. Supp., at 592. The proffered justification thus does not explain the adoption of a rule banning *99marriages by these inmates. Nor does it account for the prohibition on inmate marriages to civilians. Missouri prison officials testified that generally they had no objection to inmate-civilian marriages, see, e. g., 4 Tr. 240-241, and Superintendent Turner testified that he usually did not object to the marriage of either male or female prisoners to civilians, 2 id., at 141-142. The rehabilitation concern appears from the record to have been centered almost exclusively on female inmates marrying other inmates or ex-felons; it does not account for the ban on inmate-civilian marriages.
Moreover, although not necessary to the disposition of this case, we note that on this record the rehabilitative objective asserted to support the regulation itself is suspect. Of the several female inmates whose marriage requests were discussed by prison officials at trial, only one was refused on the basis of fostering excessive dependency. The District Court found that the Missouri prison system operated on the basis of excessive paternalism in that the proposed marriages of all female inmates were scrutinized carefully even before adoption of the current regulation — only one was approved at Renz in the period from 1979-1983 — whereas the marriages of male inmates during the same period were routinely approved. That kind of lopsided rehabilitation concern cannot provide a justification for the broad Missouri marriage rule.
It is undisputed that Missouri prison officials may regulate the time and circumstances under which the marriage ceremony itself takes place. See Brief for Respondents 5. On this record, however, the almost complete ban on the decision to marry is not reasonably related to legitimate penological objectives. We conclude, therefore, that the Missouri marriage regulation is facially invalid.
I — I <1
We uphold the facial validity of the correspondence regulation, but we conclude that the marriage rule is constitution*100ally infirm. We read petitioners’ additional challenge to the District Court’s findings of fact to be a claim that the District Court erred in holding that the correspondence regulation had been applied by prison officials in an arbitrary and capricious manner. Because the Court of Appeals did not address this question, we remand the issue to the Court of Appeals for its consideration.
Accordingly, the judgment of the Court of Appeals striking down the Missouri marriage regulation is affirmed; its judgment invalidating the correspondence rule is reversed; and the case is remanded to the Court of Appeals for further proceedings consistent with this opinion.
It is so ordered.
Suggesting that there is little difference between the “unnecessarily sweeping” standard applied by the District Court in reaching its judgment and the reasonableness standard described in Part II, see post, at 105, Justice Stevens complains that we have “ignore[d] the findings of fact that *94were made by the District Court,” post, at 102, n. 2, and have improperly “encroach[ed] into the factfinding domain of the District Court.” Post, at 101.
The District Court’s inquiry as to whether the regulations were “needlessly broad” is not just semantically different from the standard we have articulated in Part II: it is the least restrictive alternative test of Procunier v. Martinez, 416 U. S. 396 (1974). As Martinez states, in a passage quoted by the District Court:
“[T]he limitation of First Amendment freedoms must be no greater than is necessary or essential to the protection of the particular governmental interest involved. Thus a restriction on inmate correspondence . . . will ... be invalid if its sweep is unnecessarily broad.” Id., at 413-414 (emphasis added).
The District Court’s judgment that the correspondence regulation was “unnecessarily sweeping,” 586 F. Supp. 589, 596 (WD Mo. 1984), thus was a judgment based on application of an erroneous legal standard. The District Court’s findings of fact 7 and 13 likewise are predicated on application of the least restrictive means standard. Finding 7 is that the correspondence rule was applied without a letter-by-letter determination of harm, and without a showing that “there is no less restrictive alternative” available; finding 13 reiterates that the correspondence rule operated as a complete ban. See id., at 591-592. These findings are important only if petitioners have to show that the correspondence regulation satisfies a least restrictive alternative test: they are largely beside the point where the inquiry is simply whether the regulation is reasonably related to a legitimate governmental interest.
Justice Stevens’ charge of appellate factfinding likewise suffers from the flawed premise that Part III-A answers the question Justice Stevens would pose, namely, whether the correspondence regulation satisfies strict scrutiny. Thus, our conclusion that there is a logical connection between security concerns identified by petitioners and a ban on inmate-to-inmate correspondence, see supra, at 91-92, becomes, in Justice Stevens’ hands, a searching examination of the record to determine whether there was sufficient proof that inmate correspondence had actually led to an escape plot, uprising, or gang violence at Renz. See post, at 106-109. Likewise, our conclusion that monitoring inmate correspondence “clearly would impose more than a de minimis cost on the pursuit of legitimate corrections goals,” supra, at 93, is described as a factual “finding” that it *95would be “an insurmountable task” to read all correspondence sent to or received by the inmates at Renz. Post, at 110, 112. Nowhere, of course, do we make such a “finding,” nor is it necessary to do so unless one is applying a least restrictive means test.
Finally, Justice Stevens complains that Renz’ ban on inmate correspondence cannot be reasonably related to legitimate corrections goals because it is more restrictive than the rule at other Missouri institutions. As our previous decisions make clear, however, the Constitution “does not mandate a ‘lowest common denominator’ security standard, whereby a practice permitted at one penal institution must be permitted at all institutions.” Bell v. Wolfish, 441 U. S. 520, 554 (1979). Renz raises different security concerns from other Missouri institutions, both because it houses medium and maximum security prisoners in a facility without walls or guard towers, and because it is used to house inmates in protective custody. Moreover, the Renz rule is consistent with the practice of other well-run institutions, including institutions in the federal system. See Brief for United States as Amicus Curiae 22-24.