concurring in part and dissenting in part,
joined by KOZINSKI, Chief Judge, and GOULD, TALLMAN, and BEA, Circuit Judges:I agree with the majority’s conclusion regarding Byrd’s claims under the Eighth and Fourteenth Amendments. However, (1) the deference we owe prison administrators, (2) the jury’s undisputed factual findings, and (3) the relevant case law compel me to conclude that O’Connell’s actions were reasonable under the Fourth Amendment. I therefore respectfully dissent from the contradictory analysis in the majority opinion, which held that a female officer’s pat-down search of a male detainee — where there were no exigent circumstances showing the unavailability of male *1148officers to do the search — was a violation of the detainee’s Fourth Amendment rights.
In determining the reasonableness of a search under the Fourth Amendment, we must balance “the need for the particular search against the invasion of personal rights that the search entails.” Bell v. Wolfish, 441 U.S. 520, 559, 99 S.Ct. 1861, 60 L.Ed.2d 447 (1979). In Bell v. Wolfish, the Supreme Court directed us to consider the following four factors when evaluating the reasonableness of a search of an incarcerated person: (1) “the scope of the particular intrusion,” (2) “the manner in which it is conducted,” (3) “the justification for initiating it,” and (4) “the place in which it is conducted.” Id.1 The majority correctly identifies this test and properly finds that factors (3) and (4) — justification and place, respectively — weigh in favor of a finding of “reasonableness” under the Fourth Amendment. Nevertheless, it erroneously concludes that factors (1) and (2) — scope and manner, respectively — vindicate Byrd’s Fourth Amendment claim.
Before addressing the Bell factors, however, it is necessary to highlight the limited nature of our review. The Supreme Court has instructed that “[pjrison administrators ... should be accorded wide-ranging deference in the adoption and execution of policies and practices that in their judgment are needed to preserve internal order and discipline and to maintain institutional security.” Id. at 547, 99 S.Ct. 1861. *1149Therefore, “in the absence of substantial evidence in the record to indicate that the officials have exaggerated their response to[security and operation considerations], courts should ordinarily defer to their expert judgment in such matters.” Id. at 540 n. 23, 99 S.Ct. 1861 (internal quotation marks and citation omitted). It is axiomatic that prison officials know better than a panel of judges how to run a prison. Deference to prison officials, therefore, necessarily permeates our review of their searches.2
We are also constrained in our review by the jury’s undisputed factual findings that: (1) O’Connell did not “intentionally squeez[e] or knead[] [Byrd’s] penis or scrotum or improperly touch[] his anus through his underwear,” and (2) O’Connell’s search was “done for [an] identified security need.” Tempting though it may be to paint the facts differently on appeal, the jury made these findings and Byrd has not challenged them. With these considerations in mind, I now turn to the Bell factors and again emphasize that the majority agrees that two of the factors — the justification for the search and place in which it was conducted — weigh in favor of finding O’Connell’s search reasonable. Maj. Op. 1143.
It is appropriate to begin the analysis by looking at the justification for initiating the search, because, in the absence of a proper justification, even the most unintrusive search is unreasonable. The prison officials initiated their searches in response to evidence that contraband was circulating in the jail in the wake of multiple fights that had broken out in Byrd’s housing unit. This is not disputed. Maintaining the internal security of the prison was at stake, and the need to eliminate the possibility of dangerous contraband somewhere on a prisoner’s person in such a volatile atmosphere certainly justifies initiating the type of search at issue here. See Michenfelder, 860 F.2d at 333 (holding that, among other things, “testimony and physical evidence ... [of] contraband” justified the initiation of strip searches). Moreover, it is undisputed that inmates frequently attempt to conceal contraband in their body cavities, thus furthering the need for strip searches as a means of preventing the proliferation of contraband and homemade weapons within the prison. See, e.g., Bell, 441 U.S. at 559, 99 S.Ct. 1861 (“A detention facility is a unique place fraught with serious security dangers. Smuggling of money, drugs, weapons, and other contraband is all too common an occurrence. And inmate attempts to secrete these items into the facility by concealing them in body cavities are documented in this record ... and in other cases.”). On top of all this, the jury found that the search was done for an identified security need, which Byrd does not challenge on appeal. In light of such considerations, the prison’s need to confiscate all contraband, and therefore initiate these searches, was both pressing and significant. Id. at 546-47, 99 S.Ct. 1861 (“Central to all other corrections goals is the institutional consideration of internal security within the corrections facilities themselves.”) (internal quotation marks, citation, and alteration omitted).
*1150I next note my agreement with the majority that the last Bell factor — the place in which the search is conducted — also weighs in favor of finding the search reasonable. See Michenfelder, 860 F.2d at 333 (“[W]e will not question [prison officials’] judgment that conditions in [a prison unit] reasonably require searches outside the prisoners’ cells in order to protect the safety of the officers conducting them.”); see also Thompson v. Souza, 111 F.3d 694, 697, 701 (9th Cir.1997) (approving an intrusive strip search conducted in view of jeering inmates).
My agreement with the majority ends there, however, as a review of the remaining two Bell factors — the scope and manner of the search — also establishes the reasonableness of O’Connell’s search of Byrd. Relying on a string of cases that upheld both female officers’ visual observation of unclothed males and their performing pat-down searches on males, the majority concludes that, under the first Bell factor, “[t]he scope of the intrusion in this case far exceeds searches we have previously sanctioned. ...” Maj. Op. 1142. The majority is wrong.
Dealing first with the scope of O’Connell’s search, such search was limited to a pat-down of a partially clothed inmate in an attempt to locate contraband. The search was both reasonable and necessary to ensure prisoners at the Durango Jail were not concealing weapons or other contraband. Cf. Bull, 595 F.3d at 969 (“[A]rrestees’ use of body cavities as a method of smuggling drugs, weapons, and items used to escape custody is an immediate and troubling problem for San Francisco jail administrators.”). In evaluating the scope of a search, the searching officer’s gender is irrelevant.
The Supreme Court and the Ninth Circuit have previously validated even strip searches and body cavity searches. See, e.g., Bell, 441 U.S. at 558-60, 99 S.Ct. 1861 (validating visual inspections of body cavities as part of strip searches in the prison setting); Thompson, 111 F.3d 694, 700 (9th Cir.1997) (holding body cavity searches did not violate clearly established rights under qualified immunity analysis); Michenfelder, 860 F.2d at 333 (approving visual strip searches in the prison setting); Rickman v. Avaniti, 854 F.2d 327, 328 (9th Cir.1988) (approving strip searches in prisons). I recognize, of course, that these cases involved only visual searches and no touching (in contrast to O’Connell’s search). However, it is clear that pat-down searches can be constitutional even outside of a prison setting. See Terry v. Ohio, 392 U.S. 1, 17 n. 13, 88 S.Ct. 1868, 20 L.Ed.2d 889 (1968) (describing a frisk as “feelfing] with sensitive fingers every portion of the prisoner’s body ... the prisoner’s arms and armpits, waistline and back, the groin and area about the testicles, and the entire surface of the legs down to the feet” (internal quotation mark omitted)). Patdown searches are also less intrusive than strip searches. Giles v. Ackerman, 746 F.2d 614, 618 (9th Cir.1984), overruled on other grounds by Hodgers-Durgin v. de la Vina, 199 F.3d 1037 (9th Cir.1999) (en banc). The officers here searched the prisoners professionally,3 following a procedure that minimized physical contact to the extent possible. O’Connell searched Byrd over his boxer shorts, and he doesn’t claim she ever looked underneath them. The entire search took no more than a minute. Thus its scope did not “far exceed[ ]” what we have previously sanctioned. Contra Maj. Op. 1142.
*1151The final Bell factor — the manner of O’Connell’s search — gives me greater pause. Not lightly do I find reasonable a female officer’s probing search of a male detainee wearing only thin boxer shorts. Nevertheless, I believe the precedent and the facts compel this result, unsavory to our sensibilities though that result may be.
I first note “our prior case law[, which] suggests that prisoners’ legitimate expectations of bodily privacy from persons of the opposite sex are extremely limited.” Jordan v. Gardner, 986 F.2d 1521, 1524 (9th Cir.1993) (en banc). Against that backdrop, we now review our other precedent: (1) female officers may pat down the groin area of fully clothed male inmates, Grummett, 779 F.2d at 496; and (2) female officers may observe unclothed male inmates in their cells and in the showers, Michenfelder, 860 F.2d at 334. Admittedly, neither of these precedents covers the current situation — a female officer’s pat-down of a partially-clothed male inmate. Under the circumstances presented here, however, there is no meaningful difference between Byrd’s search and what we have previously upheld. O’Connell never saw Byrd’s exposed groin area or his anus, in contrast to the female officers in Michenfelder, and the search was done in a professional, swift, and appropriate manner that involved no touching of the flesh under Byrd’s boxer shorts. See Grummett, 779 F.2d at 496 (validating searches where, among other things, the searches are “done briefly ... [and] are performed by the female guards in a professional manner”). Moreover, the jury found that O’Connell did not intentionally squeeze or knead Byrd’s penis or scrotum or improperly touch his anus through the boxer shorts and that O’Connell’s search was done for an identified security need. Again, Byrd has not challenged these findings.
In addition, the County has offered undisputed evidence that the County did not have sufficient numbers of male detention officers to conduct searches of male inmates without the assistance of female officers.4 Thus, even if it may seem odd to utilize a female officer to conduct a pat-down of a male inmate when other male officers are present, the realities of the prison’s staffing needs justify such searches. See Michenfelder, 860 F.2d at 334 (“[R]equiring [female employees] to be replaced by males for the duration of strip searches, would displace officers through*1152out the prison.”); Grummett, 779 F.2d at 496 (“To restrict the female guards from positions which involve occasional viewing of the inmates would necessitate a tremendous rearrangement of work schedules, and possibly produce a risk to both internal security needs and equal employment opportunities for the female guards.”). This is particularly true since the 10,000 detainees in the Maricopa County system are searched at least daily, and often several times a day. To disallow all female officers from conducting these searches anytime a male officer is present would handicap the female officers as security personnel and perpetuate sexist notions that a female is only useful when a male is not available. Instead of converting Durango Jail into a target for equal employment litigation, I defer to the prison officials’ reasoned and sensible judgment on these matters.
In order to provide a basis for its decision, the majority terms this search a strip search, rather than a pat-down search. This is error. True, this search differed somewhat from the prison’s definition of a “frisk (body) search,” because more than Byrd’s socks and shoes were removed. However, it is undisputed that it involved only an inspection of Byrd’s (already removed) clothing and “feeling the contours of his clothed body.” O’Connell testified that at no point did she pat down or touch any unclothed areas. Thus, the search is more accurately described as a “pat-down,” since a “strip search” involves a “visual scan of the inmate’s body after all clothing has been removed.” Indeed, the strip searches in the cases relied on by the majority were visual inspections of naked inmates. See Moore v. Carwell, 168 F.3d 234, 236 (5th Cir.1999) (male inmate “being viewed naked by a female [officer]”); Canedy v. Boardman, 16 F.3d 183, 185, 186 n. 2 (7th Cir.1994) (female guards’ visual observations of male inmate’s “naked body” prohibited, but pat-down searches constitutional); Hayes v. Marriott, 70 F.3d 1144, 1145-47 (10th Cir.1995) (Visual body cavity search of male inmate in presence of nonessential female staff. The court noted “the Fourth Amendment does not require the complete exclusion of members of the opposite sex from areas in which searches are conducted”); Cookish v. Powell, 945 F.2d 441, 444-45 (1st Cir.1991) (male inmate “had to remove every item of clothing” in view of supervising female officer); Lee v. Downs, 641 F.2d 1117, 1120 (4th Cir.1981) (female inmate had “underclothing ... forcefully removed” with male guards present); see also Sec. and Law Enforcement Employees, Dist. Council 82 v. Carey, 737 F.2d 187, 192 nn.3-4 (2d Cir.1984) (distinguishing between “strip frisks” (requiring the person being searched to reveal body cavities), visual examinations of naked persons, and pat frisks); Cookish, 945 F.2d at 444 n. 5 (“A ‘strip search,’ ... refers to an inspection of a naked individual”). None of these “strip search” cases involved partially clothed inmates.
The majority’s description of this search as a strip search allows it to discount this court’s precedent relevant to the case at hand. Grummett is on point. This court held that “routine pat-down searches, which include the groin area, and which are otherwise justified by security needs, do not violate the [Fourteenth Amendment because a correctional officer of the opposite gender conducts such a search.” 779 F.2d at 495; see id. at 496 (finding no Fourth Amendment violation). The majority repeatedly distinguishes Grummett because it feels that, unlike Grummett, this case involved “intimate contact with the inmate’s body.” Maj. Op. 1143 n.8. This, however, ignores the nature and purpose of a pat-down search. A pat-down is done to detect contraband that may be taped to the contours of an inmate’s body, including the genital area. One has to *1153apply enough pressure to contact the areas being searched, and to “be able to feel something that’s on the [area searched].” Given that the purpose of the search is to feel for contraband, the majority’s attempt to distinguish Grummetb solely on the thickness or extent of the clothing is unconvincing. In order to be effective, the searcher must be able to feel the inside thigh and perineum. On an inmate in street clothes, this may require more pressure, but results in the same contact in the same area. In both cases, no skin to skin contact was made; both searches were only over clothed areas. Both searches were of the same groin area. Both searches were brief and conducted for legitimate security purposes. 779 F.2d at 496. Given the area, nature, and purpose of the search, it would be unreasonable to presume that the female guards in Grummetb did not face the same challenges O’Connell did.
Moreover, even if this were a strip search, this court’s precedent does not suggest that a cross-gender strip search is unconstitutional where the genitals remain covered throughout the procedure. After considering Grummetb, Jordan, and Michenfelder, this court noted that “it is highly questionable even today whether prison inmates have a Fourth Amendment right to be free from routine unclothed searches by officials of the opposite sex, or from viewing of their unclothed bodies by officials of the opposite sex.” Somers v. Thurman, 109 F.3d 614, 622 (9th Cir.1997). This is even more true when the special concern about “involuntary exposure of [one’s genitals] in the presence of people of the other sex” is not implicated. Fortner v. Thomas, 983 F.2d 1024, 1030 (11th Cir.1993).
Lastly, the majority seems to rest its analysis of the “manner” factor on the fact that O’Connell was unidentified to Byrd and was wearing jeans and a t-shirt and that there were ten to fifteen non-participating officers present. Maj. Op. 1142-43. Even putting aside our heightened deference to prison officials, the majority does not explain how the number of officers, their attire, or their lack of identification somehow outweigh the jury’s finding that the search was appropriate,5 the professional nature of the search, or the County’s undisputed need to utilize its female officers (or even how these facts are relevant). Nor were such facts unreasonable in any event. O’Connell testified that Byrd was allowed to ask her questions but didn’t. He talked to other officers after the search was over, and they explained that cadets had helped conduct the search. It was also not unreasonable for some officers not to participate in the search. We see no reason to intimate that a search is constitutionally suspect because not every single *1154officer present laid hands on the prisoner. Surely this cuts the other way, if it cuts at all.6
The majority’s holding on this point illustrates the risks of not giving proper deference to prison officials. As part of the prison’s search for contraband, officials ordered more than ninety inmates out of their cells into the common area. In contrast, there were only twenty-five to thirty Academy cadets and ten to fifteen uniformed detention officers present. At best, the prison officials were outnumbered two to one. Importantly, judges are instructed not to intrude into that sphere of decision making. Deciding to accelerate the searches by using additional assistance from female cadets hardly seems unreasonable in these circumstances. I question second-guessing prison administrators when such credible threats to officers’ safety are present. Indeed; I would think such security concerns vindicate the Supreme Court’s instruction to defer to prison administrators — an instruction that it seems the majority has too hastily disregarded.
In sum, we are guided by our precedent and bound by Supreme Court precedent, the jury’s undisputed factual findings, and our deference to prison officials’ expertise in these matters. Balancing the four Bell factors in light of these limitations, I conclude that O’Connell’s search of Byrd was reasonable under the Fourth Amendment, and I respectfully dissent from the contradictory analysis in the majority opinion.
. Because the majority holds that the factors articulated in Turner v. Safley, 482 U.S. 78, 89-91, 107 S.Ct. 2254, 96 L.Ed.2d 64 (1987), are inapplicable to Byrd’s challenge and instead focuses exclusively on the Bell factors, Maj. Op. 1141 n.6, I too only address the Bell factors. It ultimately makes no difference whether we consider the factors articulated in Turner, however, as application of the Turner factors also compels a finding that O'Connell's search of Byrd was reasonable under the Fourth Amendment. See Bull v. City and County of San Francisco, 595 F.3d 964, 975-76 (9th Cir.2010) (en banc) ("Because the Turner factors require us to give more deference to detention officials' determinations than does the balancing test in Bell, it is not surprising that our consideration of the Turner factors leads to the same conclusion.”). In Turner, the Supreme Court held that "when a prison regulation impinges on inmates’ constitutional rights, the regulation is valid if it is reasonably related to legitimate penological interests.” 482 U.S. at 89, 107 S.Ct. 2254. In making this determination, the Supreme Court directed us to consider four factors. Id. at 89-91, 107 S.Ct. 2254. First, we consider whether there is a "valid, rational connection between the prison regulation and the legitimate governmental interest put forward to justify it.” Id. at 89, 107 S.Ct. 2254 (internal quotation marks and citation omitted). As discussed more fully infra, it is not disputed that the prison officials had a legitimate need to initiate the searches in order to prevent the proliferation of contraband within the prison. Second, we are to consider "whether there are alternative means of exercising the right that remain open to prison inmates.” Id. at 90, 107 S.Ct. 2254. This factor does not apply to Byrd's Fourth Amendment claim; it is somewhat nonsensical to examine whether alternative means exist for Byrd to exercise his right to be free from unreasonable searches. See Michenfelder v. Sumner, 860 F.2d 328, 331 n. 1 (9th Cir.1988) ("Not all four factors will be relevant to each case.... [T]he second Turner factor ... is much more meaningful in the first amendment context than the fourth or eighth....”). Third, we consider "the impact accommodation of the asserted constitutional right will have on guards and other inmates, and on the allocation of prison resources generally.” Turner, 482 U.S. at 90, 107 S.Ct. 2254. Again as discussed more fully infra, the County offered undisputed evidence that it did not have sufficient numbers of male detention officers to conduct searches of male inmates without the assistance of female officers. Finally, we must examine "the absence of ready alternatives [as] evidence of the reasonableness of a prison regulation.” Id. Because we must defer to the County's undisputed evidence regarding its staffing needs, we cannot conclude that there were other reasonable alternatives to the search.
. The majority has apparently disregarded this principle, instead relying on a Commission Report and two op-eds to form its own opinions about the logistical wisdom of the searches at issue here. Maj. Op. 1140-41. Aside from the fact that the Commission Report and op-eds are not part of the record before us, there is certainly nothing to suggest that they represent the collective wisdom of administrators who understand the conditions at Durango Jail. Those prison administrators deserve greater deference, especially since the evaluation of the reasonableness of a search is heavily dependent on the context of each case.
. The majority’s assertion that it makes no difference that a search was done in a professional manner is unconvincing. Maj. Op. 1142-43, 1144-45. The court in Grummett specifically noted (three times) that female guards’ searches were done in “in a professional manner.” Grummett v. Rushen, 779 F.2d 491, 495-96 (9th Cir.1985).
. The majority argues that Byrd disputed the need to utilize female officers because he testified that male detention officers were present, but did not conduct the search. Maj. Op. 1137 & n.2. The majority claims that, as in Moore v. Carwell, 168 F.3d 234, 237 (5th Cir.1999), we should credit Byrd’s testimony that some male officers were not participating in the search as evidence that female officers were not needed to perform the search. However, in Moore, the court was required to assume Moore's allegations were true when reviewing a motion to dismiss. Id. at 236. We are not so constrained in our review; instead, we must consider all the evidence produced at trial. While Byrd testified that not all male officers present conducted the searches, he did not (and could not) testify as to the staffing necessary to safely conduct a search of approximately 90 prisoners and their living quarters. Moreover, the majority's argument that female officers were not necessary for the search was contradicted by testimony from both O’Connell and Captain Peterson. O’Connell testified that having additional officers present was necessary because ”[t]here is always a threat” of violence in such situations. Captain Peterson testified that prison officials "don’t have the luxury” of having enough male detention officers to search the jail facility. Therefore, they must rely on female officers. He also testified that he could not pull in additional male officers from other facilities to have enough male staff to conduct such searches. Under Bell, we are required to defer to the expert judgment of prison officials as to how to maintain prison safety and order — including who should conduct searches and how many officers are needed.
. The majority explains that this is relevant because of the "humiliation” of "exposing one’s nude body to strangers.” Maj. Op. 1136 n.l. However, it does not explain how identification or uniforms would make the cadets not "strangers.” Given their matching white shirts, supervision by other guards, and participation in the search, it was surely apparent that the cadets were present in an official capacity, not just strangers off the street. Moreover, contrary to the majority’s assertion, Byrd was not nude. Maj. Op. 1136, 383 nn.l, 8. There is no basis in the record for its assertion that the pink boxers were sheer, see-through, or even nearly see-through (which really means not see-through). The majority’s reference to Arizona and Hawaii law for the proposition that appearing in public dressed in boxer shorts would be illegal is also unfounded. First, the laws set standards for obscenity, not indecent exposure. Second, and more importantly, the record is clear that at no point did O’Connell see any area covered by the boxer shorts, including the genital area. Therefore, the oft-quoted "desire to shield one’s unclothed figure from the view of strangers,” York v. Story, 324 F.2d 450, 455 (9th Cir.1963), was satisfied in this case.
. The majority also relies on the fact that one or two people in the room, about 30 feet away from Byrd, had video cameras. Maj. Op. 1142-43. But we have previously found that some strip searches were reasonable even when transmitted to video screens monitored by guards of the opposite sex. Michenfelder, 860 F.2d at 329-30, 333.