delivered the opinion of the Court.
In this case, we must decide whether a state hospital’s performance of a diagnostic test to obtain evidence of a patient’s criminal conduct for law enforcement purposes is an *70unreasonable search if the patient has not consented to the procedure. More narrowly, the question is whether the interest in using the threat of criminal sanctions to deter pregnant women from using cocaine can justify a departure from the general rule that an official nonconsensual search is unconstitutional if not authorized by a valid warrant.
I
In the fall of 1988, staff members at the public hospital operated in the city of Charleston by the Medical University of South Carolina (MUSC) became concerned about an apparent increase in the use of cocaine by patients who were receiving prenatal treatment.1 In response to this perceived increase, as of April 1989, MUSC began to order drug screens to be performed on urine samples from maternity patients who were suspected of using cocaine. If a patient tested positive, she was then referred by MUSC staff to the county substance abuse commission for counseling and treatment. However, despite the referrals, the incidence of cocaine use among the patients at MUSC did not appear to change.
Some four months later, Nurse Shirley Brown, the case manager for the MUSC obstetrics department, heard a news broadcast reporting that the police in Greenville, South Carolina, were arresting pregnant users of cocaine on the theory that such use harmed the fetus and was therefore child abuse.2 Nurse Brown discussed the story with MUSC’s general counsel, Joseph C. Good, Jr., who then contacted *71Charleston Solicitor Charles Condon in order to offer MUSC’s cooperation in prosecuting mothers whose children tested positive for drugs at birth.3
After receiving Good’s letter, Solicitor Condon took the first steps in developing the policy at issue in this case. He organized the initial meetings, decided who would participate, and issued the invitations, in which he described his plan to prosecute women who tested positive for cocaine while pregnant. The task force that Condon formed included representatives of MUSC, the police, the Gounty Substance Abuse Commission and the Department of Social Services. Their deliberations led to MUSC’s adoption of a 12-page document entitled “POLICY M-7,” dealing with the subject of “Management of Drug Abuse During Pregnancy.” App. to Pet. for Cert. A-53.
The first three pages of Policy M-7 set forth the procedure to be followed by the hospital staff to “identify/assist pregnant patients suspected of drug abuse.” Id., at A-53 to A-56. The first section, entitled the “Identification of Drug Abusers,” provided that a patient should be tested for co-eaine through a urine drug screen if she met one or more of nine criteria.4 It also stated that a chain of custody should *72be followed when obtaining and testing urine samples, presumably to make sure that the results could be used in subsequent criminal proceedings. The policy also provided for education and referral to a substance abuse clinic for patients who tested positive. Most important, it added the threat of law enforcement intervention that “provided the necessary ‘leverage’ to make the [p]oliey effective.” Brief for Respondents 8. That threat was, as respondents candidly acknowledge, essential to the program’s success in getting women into treatment and keeping them there.
The threat of law enforcement involvement was set forth in two protocols, the first dealing with the identification of drug use during pregnancy, and the seeond with identification of drug use after labor. Under the latter protocol, the police were to be notified without delay and the patient promptly arrested. Under the former, after the initial positive drug test, the police were to be notified (and the patient arrested) only if the patient tested positive for cocaine a see-ond time or if she missed an appointment with a substance abuse counselor.5 In 1990, however, the policy was modified at the behest of the solicitor’s office to give the patient who tested positive during labor, like the patient who tested positive during a prenatal care visit, an opportunity to avoid arrest by consenting to substance abuse treatment.
The last six pages of the policy contained forms for the patients to sign, as well as procedures for the police to follow when a patient was arrested. The policy also prescribed in detail the precise offenses with which a woman could be charged, depending on the stage of her pregnancy. If the pregnancy was 27 weeks or less, the patient was to be charged with simple possession. If it was 28 weeks or more, she was to be charged with possession and distribution to a person under the age of 18 — in this case, the fetus. If she *73delivered "while testing positive for illegal drugs,” she was also to be charged with unlawful neglect of a child. App. to Pet. for Cert. A-62. Under the policy, the police were instructed to interrogate the arrestee in order "to ascertain the identity of the subject who provided illegal drugs to the suspect.” Id., at A-63. Other than the provisions describing the substance abuse treatment to be offered to women who tested positive, the policy made no mention of any change in the prenatal care of such patients, nor did it prescribe any special treatment for the newborns.
HH
Petitioners are 10 women who received obstetrical care at MUSC and who were arrested after testing positive for cocaine. Four of them were arrested during the initial implementation of the policy; they were not offered the opportunity to receive drug treatment as an alternative to arrest. The others were arrested after the policy was modified in 1990; they either failed to comply with the terms of the drug treatment program or tested positive for a second time. Respondents include the city of Charleston, law enforcement officials who helped develop and enforce the policy, and representatives of MUSC.
Petitioners’ complaint challenged the validity of the policy under various theories, including the claim that warrantless and nonconsensual drug tests conducted for criminal investigatory purposes were unconstitutional searches. Respondents advanced two principal defenses to the constitutional claim: (1) that, as a matter of fact, petitioners had consented to the searches; and (2) that, as a matter of law, the searches were reasonable, even absent consent, because they were justified by special non-law-enforcement purposes. The District Court rejected the second defense because the searches in question “were not done by the medical university for independent purposes. [Instead,] the police came in and there was an agreement reached that the positive *74screens would be shared with the police.” App. 1248-1249. Accordingly, the District Court submitted the factual defense to the jury with instructions that required a verdict in favor of petitioners unless the jury found consent.6 The jury found for respondents.
Petitioners appealed, arguing that the evidence was not sufficient to support the jury’s consent finding. The Court of Appeals for the Fourth Circuit affirmed, but without reaching the question of consent. 186 F. 3d 469 (1999). Disagreeing with the District Court, the majority of the appellate panel held that the searches were reasonable as a matter of law under our line of cases recognizing that “special needs” may, in certain exceptional circumstances, justify a search policy designed to serve non-law-enforcement ends.7 *75On the understanding “that MUSC personnel conducted the urine drug screens for medical purposes wholly independent of an intent to aid law enforcement efforts,”8 id., at 477, the majority applied the balancing test used in Treasury Employees v. Von Raab, 489 U. S. 656 (1989), and Vernonia School Dist. 47J v. Acton, 515 U. S. 646 (1995), and concluded that the interest in curtailing the pregnancy complications and medical costs associated with maternal cocaine use outweighed what the majority termed a minimal intrusion on the privacy of the patients. In dissent, Judge Blake concluded that the “special needs” doctrine should not apply and *76that the evidence of consent was insufficient to sustain the jury’s verdict. 186 F. 3d, at 487-488.
We granted certiorari, 528 U.S. 1187 (2000), to review the appellate court’s holding on the “special needs” issue. Because we do not reach the question of the sufficiency of the evidence with respect to consent, we necessarily assume for purposes of our decision — as did the Court of Appeals — that the searches were conducted without the informed consent of the patients. We conclude that the judgment should be reversed and the ease remanded for a decision on the consent issue.
Ill
Because MUSO is a state hospital, the members of its staff are government actors, subject to the strictures of the Fourth Amendment. New Jersey v. T. L. O., 469 U. S. 325, 335-337 (1985). Moreover, the urine tests conducted by those staff members were indisputably searches within the meaning of the Fourth Amendment. Skinner v. Railway Labor Executives’ Assn., 489 U. S. 602, 617 (1989).9 Neither the District Court nor the Court of Appeals concluded that any of the nine criteria used to identify the women to be searched provided either probable cause to believe that they were using cocaine, or even the basis for a reasonable suspicion of such use. Rather, the District Court and the Court of Appeals viewed the case as one involving MUSC’s right *77to conduct searches -without warrants or probable cause.10 Furthermore, given the posture in which the case comes to us, we must assume for purposes of our decision that the tests were performed without the informed consent of the patients.11
Because the hospital seeks to justify its authority to conduct drug tests and to turn the results over to law enforcement agents without the knowledge or consent of the patients, this case differs from the four previous cases in which we have considered whether comparable drug tests “fit within the closely guarded category of constitutionally permissible suspicionless searches.” Chandler v. Miller, 520 U. S. 305, 309 (1997). In three of those cases, we sustained drug tests for railway employees involved in train accidents, Skinner v. Railway Labor Executives’ Assn., 489 U. S. 602 (1989), for United States Customs Service employees seeking promotion to certain sensitive positions, Treasury Employees v. Von Raab, 489 U. S. 656 (1989), and for high school students participating in interseholastic sports, Vernonia School Dist. 47J v. Acton, 515 U. S. 646 (1995). In the fourth case, we struck down such testing for candidates for designated state offices as unreasonable. Chandler v. Miller, 520 U. S. 305 (1997).
*78In each of those cases, we employed a balancing test that weighed the intrusion on the individual’s interest in privacy against the “special needs” that supported the program. As an initial matter, we note that the invasion of privacy in this case is far more substantial than in those eases. In the previous four cases, there was no misunderstanding about the purpose of the test or the potential use of the test results, and there were protections against the dissemination of the results to third parties.12 The use of an adverse test result to disqualify one from eligibility for a particular benefit, such as a promotion or an opportunity to participate in an extracurricular activity, involves a less serious intrusion on privacy than the unauthorized dissemination of such results to third parties. The reasonable expectation of privacy enjoyed by the typical patient undergoing diagnostic tests in a hospital is that the results of those tests will not be shared with nonmedical personnel without her consent. See Brief for American Medical Association as Amicus Curiae 11; Brief for American Public Health Association et al. as Amici Curiae 6,17-19.13 In none of our prior cases was there any intrusion upon that kind of expectation.14
*79The critical difference between those four drug-testing cases and this one, however, lies in the nature of the “special need” asserted as justification for the warrantless searches. In each of those earlier cases, the “special need” that was advanced as a justification for the absence of a warrant or individualized suspicion was one divorced from the State’s general interest in law enforcement.15 This point was em*80phasized both in the majority opinions sustaining the programs in the first three eases,16 as well as in the dissent in the Chandler ease.17 In this case, however, the central and indispensable feature of the policy from its inception was the use of law enforcement to coerce the patients into substance abuse treatment. This fact distinguishes this case from circumstances in which physicians or psychologists, in the *81course of ordinary medical procedures aimed at helping the patient herself, come across information that under rules of law or ethics is subject to reporting requirements, which no one has challenged here. See, e. g., Council on Ethical and Judicial Affairs, American Medical Association, Policy-Finder, Current Opinions E-5.05 (2000) (requiring reporting where “a patient threatens to inflict serious bodily harm to another person or to him or herself and there is a reasonable probability that the patient may carry out the threat”); Ark. Code Ann. § 12-12-602 (1999) (requiring reporting of intentionally inflicted knife or gunshot wounds); Ariz. Rev. Stat. Ann. § 13-3620 (Supp. 2000) (requiring “any ... person having responsibility for the care or treatment of children” to report suspected abuse or neglect to a peace officer or child protection agency).18
Respondents argue in essence that their ultimate purpose — namely, protecting the health of both mother and child — is a beneficent one. In Chandler, however, we did not simply accept the State’s invocation of a “special need.” Instead, we carried out a “close review” of the scheme at issue before concluding that the need in question was not “special,” as that term has been defined in our cases. 520 U.S., at 322. In this case, a review of the M-7 policy plainly reveals that the purpose actually served by the MUSC searches “is ultimately indistinguishable from the general interest in crime control.” Indianapolis v. Edmond, 531 U.S. 32, 44 (2000).
In looking to the programmatic purpose, we consider all the available evidence in order to determine the relevant primary purpose. See, e. g., id., at 45-47. In this case, as *82Judge Blake put it in her dissent below, “it ... is clear from the record that an initial and continuing focus of the policy was on the arrest and prosecution of drug-abusing mothers . . . .” 186 F. 3d, at 484. Tellingly, the document codifying the policy incorporates the police’s operational guidelines. It devotes its attention to the chain of custody, the range of possible criminal charges, and the logistics of police notification and arrests. Nowhere, however, does the document discuss different courses of medical treatment for either mother or infant, aside from treatment for the mother’s addiction.
Moreover, throughout the development and application of the policy, the Charleston prosecutors and police were extensively involved in the day-to-day administration of the policy. Police and prosecutors decided who would receive the reports of positive drug screens and what information would be included with those reports. App. 78-80,145-146,1058-1060. Law enforcement officials also helped determine the procedures to be followed when performing the screens.19 Id., at 1052-1053. See also id., at 26-27,945. In the course of the policy’s administration, they had access to Nurse Brown’s medical files on the women who tested positive, routinely attended the substance abuse team’s meetings, and regularly received copies of team documents discussing the women’s progress. Id., at 122-124, 609-610. Police took pains to coordinate the timing and circumstances of the arrests with MUSC staff, and, in particular, Nurse Brown. Id., at 1057-1058.
While the ultimate goal of the program may well have been to get the women in question into substance abuse treatment *83and off of drugs, the immediate objective of the searches was to generate evidence for law enforcement purposes20 in order to reach that goal.21 The threat of law enforcement *84may ultimately have been intended as a means to an end, but the direct and primary purpose of MUSC’s policy was to ensure the use of those means. In our opinion, this distinction is critical. Because law enforcement involvement always serves some broader social purpose or objective, under respondents’ view, virtually any nonconsensual suspicionless search could be immunized under the special needs doctrine by defining the search solely in terms of its ultimate, rather than immediate, purpose.22 Such an approach is inconsistent with the Fourth Amendment. Given the primary purpose of the Charleston program, which was to use the threat of arrest and prosecution in order to force women into treatment, and given the extensive involvement of law enforcement officials at every stage of the policy, this case simply does not fit within the closely guarded category of "special needs.”23
The fact that positive test results were turned over to the police does not merely provide a basis for distinguishing our prior cases applying the “special needs” balancing approach to the determination of drug use. It also provides an affirmative reason for enforcing the strictures of the Fourth Amendment. While state hospital employees, like other citizens, may have a duty to provide the police with evidence *85of criminal conduct that they inadvertently acquire in the course of routine treatment, when they undertake to obtain such evidence from their patients for the specific purpose of incriminating those patients, they have a special obligation to make sure that the patients are fully informed about their constitutional rights, as standards of knowing waiver require.24 Cf. Miranda v. Arizona, 384 U.S. 436 (1966).
As respondents have repeatedly insisted, their motive was benign rather than punitive. Such a motive, however, cannot justify a departure from Fourth Amendment protections, given the pervasive involvement of law enforcement with the development and application of the MUSC policy. The stark *86and unique fact that characterizes this case is that Policy M-7 was designed to obtain evidence of criminal conduct by the tested patients that would be turned over to the police and that could be admissible in subsequent criminal prosecutions. While respondents are correct that drug abuse both was and is a serious problem, “the gravity of the threat alone cannot be dispositive of questions concerning what means law enforcement officers may employ to pursue a given purpose.” Indianapolis v. Edmond, 531 U. S., at 42-43. The Fourth Amendment’s general prohibition against nonconsen-sual, warrantless, and suspicionless searches necessarily applies to such a policy. See, e. g., Chandler, 520 U. S., at 308; Skinner, 489 U. S., at 619.
Accordingly, the judgment of the Court of Appeals is reversed, and the case is remanded for further proceedings consistent with this opinion.
It is so ordered.
As several witnesses testified at trial, the problem of “crack babies” was widely perceived in the late 1980’s as a national epidemic, prompting considerable concern both in the medical community and among the general populace.
Under South Carolina law, a viable fetus has historically been regarded as a person; in 1995, the South Carolina Supreme Court held that the ingestion of cocaine during the third trimester of pregnancy constitutes criminal child neglect. Whitner v. South Carolina, 328 S. C. 1,492 S. E. 2d 777 (1995), cert. denied, 523 U. S. 1145 (1998).
In his letter dated August 23, 1989, Good wrote: “Please advise us if your office is anticipating future criminal action and what if anything our Medical Center needs to do to assist you in this matter.” App. to Pet. for Cert. A-67.
Those criteria were as follows:
“1. No prenatal care
“2. Late prenatal care after 24 weeks gestation
“3. Incomplete prenatal care
“4. Abruptio placentae
“5. Intrauterine fetal death
“6. Preterm labor ‘of no obvious cause’
“7. IUGR [intrauterine growth retardation] ‘of no obvious cause’
“8. Previously known drug or alcohol abuse
“9. Unexplained congenital anomalies.” Id., at A-53 to A-54.
Despite the conditional description of the first category, when the policy was in its initial stages, a positive test was immediately reported to the police, who then promptly arrested the patient.
The instructions read: “THERE WERE NO SEARCH WARRANTS ISSUED BY A MAGISTRATE OR ANY OTHER PROPER JUDICIAL OFFICER TO PERMIT THESE URINE SCREENS TO BE TAKEN. THERE NOT BEING A WARRANT ISSUED, THEY ARE UNREASONABLE AND IN VIOLATION OF THE CONSTITUTION OF THE UNITED STATES, UNLESS THE DEFENDANTS HAVE SHOWN BY THE GREATER WEIGHT OR PREPONDERANCE OF THE EVIDENCE THAT THE PLAINTIFFS CONSENTED TO THOSE SEARCHES.” App. 1314r-1315. Under the judge’s instructions, in order to find that the plaintiffs had consented to the searches, it was necessary for the jury to find that they had consented to the taking of the samples, to the testing for evidence of cocaine, and to the possible disclosure of the test results to the police. Respondents have not argued, as Justice Scaua does, that it is permissible for members of the staff of a public hospital to use diagnostic tests “deceivingly” to obtain incriminating evidence from their patients. See post, at 94 (dissenting opinion).
The term “special needs” first appeared in Justice Blackmun’s opinion concurring in the judgment in New Jersey v. T. L. O., 469 U. S. 325, 351 (1985). In his concurrence, Justice Blackmun agreed with the Court that there are limited exceptions to the probable-cause requirement, in which reasonableness is determined by “a careful balancing of governmental and private interests,” but concluded that such a test should only be applied ^‘in those exceptional circumstances in which special needs, beyond the normal need for law enforcement, make the warrant and probable-cause requirement impracticable_.” Ibid. This Court subsequently *75adopted the “special needs” terminology in O’Connor v. Ortega, 480 U.S. 709, 720 (1987) (plurality opinion), and Griffin v. Wisconsin, 483 U. S. 868, 873 (1987), concluding that, in limited circumstances, a search unsupported by either warrant or probable cause can be constitutional when “special needs” other than the normal need for law enforcement provide sufficient justification. See also Vernonia School Dist. 47J v. Acton, 515 U.S. 646, 652-653 (1995).
The majority stated that the District Court had made such a finding. 186 F. 3d, at 477. The text of the relevant finding, made in the context of petitioners’ now abandoned Title VI claim, reads as follows: “The policy was applied in all maternity departments at MUSC. Its goal was not to arrest patients but to facilitate their treatment and protect both the mother and unborn child,” App. to Pet. for Cert. A-38. That finding, however, must be read in light of this comment by the District Court with respect to the Fourth Amendment claim:
“. . . THESE SEARCHES WERE NOT DONE BY THE MEDICAL UNIVERSITY FOR INDEPENDENT PURPOSES. IF THEY HAD BEEN, THEN THEY WOULD NOT IMPLICATE THE FOURTH AMENDMENT. OBVIOUSLY AS I POINT OUT THERE ON PAGE 4, NORMALLY URINE SCREENS AND BLOOD TESTS AND THAT TYPE OF THING CAN BE TAKEN BY HEALTH CARE PROVIDERS WITHOUT HAVING TO WORRY ABOUT THE FOURTH AMENDMENT. THE ONLY REASON THE FOURTH AMENDMENT IS IMPLICATED HERE IS THAT THE POLICE CAME IN AND THERE WAS AN AGREEMENT REACHED THAT THE POSITIVE SCREENS WOULD BE SHARED WITH THE POLICE. AND THEN THE SCREEN IS NOT DONE INDEPENDENT OF POLICE, IT’S DONE IN CONJUNCTION WITH THE POLICE AND THAT IMPLICATES THE FOURTH AMENDMENT.” App. 1248-1249.
In arguing that the urine tests at issue were not searches, the dissent attempts to disaggregate the taking and testing of the urine sample from the reporting of the results to the police. See post, at 92. However, in our special needs cases, we have routinely treated mine screens taken by state agents as searches within the meaning of the Fourth Amendment even though the results were not reported to the police, see, e. g., Chandler v. Miller, 520 U. S. 305 (1997); Vernonia School Dish 4J v. Acton, 515 U. S. 646 (1995); Skinner v. Railway Labor Executives’ Assn., 489 U. S. 602, 617 (1989); Treasury Employees v. Von Raab, 489 U. S. 656 (1989), and respondents here do not contend that the tests were not searches. Rather, they argue that the searches were justified by consent and/or by special needs.
In a footnote to their brief, respondents do argue that the searches were not entirely suspicionless. Brief for Respondents 23, n. 13. They do not, however, point to any evidence in the record indicating that any of the nine search criteria was more apt to be caused by cocaine use than by some other factor, such as malnutrition, illness, or indigency. More significantly, their legal argument and the reasoning of the majority panel opinion rest on the premise that the policy would be valid even if the tests were conducted randomly.
The dissent would have us do otherwise and resolve the issue of consent in favor of respondents. Because the Court of Appeals did not discuss this issue, we think it more prudent to allow that court to resolve the legal and factual issues in the first instance, and we express no view on those issues. See, e. g., Glover v. United States, 531 U. S. 198 (2001); National Collegiate Athletic Assn. v. Smith, 525 U.S. 459, 470 (1999).
Chandler, 520 U. S., at 312, 318; Acton, 515 U. S., at 658; Skinner, 489 U. S., at 621, n. 5,622, n. 6; Von Saab, 489 U. S., at 663,666-667, 672, n. 2.
There are some circumstances in which state hospital employees, like other citizens, may have a duly to provide law enforcement officials with evidence of criminal conduct acquired in the course of routine treatment, see, e. g., S. G. Code Ann. §20-7-510 (2000) (physicians and nurses required to report to child welfare ageney or law enforcement authority “when in the person’s professional capacity the person” receives information that a child has been abused or neglected). While the existence of such laws might lead a patient to expect that members of the hospital staff might turn over evidence acquired in the course of treatment to which the patient had consented, they surely would not lead a patient to anticipate that hospital staff would intentionally set out to obtain incriminating evidence from their patients for law enforcement purposes.
In fact, we have previously recognized that an intrusion on that expectation may have adverse consequences because it may deter patients from *79receiving needed medical care. Whalen v. Roe, 429 U. S. 589, 599-600 (1977). Cf. Poland, Dombrowski, Ager, & Sokol, Punishing pregnant drug users: enhancing the flight from care, 31 Drug and Alcohol Dependence 199-203 (1993).
As The Chief Justice recently noted: “The ‘special needs’ doctrine, which has been used to uphold certain suspicionless searches performed for reasons unrelated to law enforcement, is an exception to the general rule that a search must be based on individualized suspicion of wrongdoing.” Indianapolis v. Edmond, 531 U.S. 32, 54 (2000) (dissenting opinion); see also nn. 16-17, infra. TnT.L. O., we made a point of distinguishing searches “carried out by school authorities acting alone and on their own authority” from those conducted “in conjunction with or at the behest of law enforcement agencies.” 469 U. S., at 341, n. 7.
The dissent, however, relying on Griffin v. Wisconsin, 483 U.S. 868 (1987), argues that the special needs doctrine “is ordinarily employe[d], precisely to enable searches by law enforcement officials who, of course, ordinarily have a law enforcement objective.” Post, at 100. Viewed in the context of our special needs case law and even viewed in isolation, Griffin does not support the proposition for which the dissent invokes it. In other special needs cases, we have tolerated suspension of the Fourth Amendment’s warrant or probable-cause requirement in part because there was no law enforcement purpose behind the searches in those cases, and there was little, if any, entanglement with law enforcement. See Skinner, 489 U. S., at 620-621; Von Raab, 489 U.S., at 665-666; Acton, 515 U.S., at 658. Moreover, after our decision in Griffin, we reserved the question whether “routine use in criminal prosecutions of evidence obtained pursuant to the administrative scheme would give rise to an inference of pretext, or otherwise impugn the administrative nature of the... program.” Skinner, 489 U. S., at 621, n. 5. In Griffin itself, this Court noted that “[although a probation officer is not an impartial magistrate, neither is he the police officer who normally conducts searches against the ordinary citizen.” 483 U.S., at 876. Finally, we agree with petitioners *80that Griffin is properly read as limited by the fact that probationers have a lesser expectation of privacy than the public at large. Id,., at 874-875.
In Skinner v. Railway Labor Executives’ Assn., 489 U. S. 602 (1989), this Court noted that “[t]he FRA has prescribed toxicological tests, not to assist in the prosecution of employees, but rather ‘to prevent accidents and casualties in railroad operations that result from impairment of employees by alcohol or drugs.’ ” Id., at 620-621 (quoting 49 CFR § 219.1(a) (1987)). Similarly, in Treasury Employees v. Von Raab, 489 U. S. 656 (1989), we concluded that it was "clear that the Customs Service’s drug-testing program is not designed to serve the ordinary needs of law enforcement. Test results may not be used in a criminal prosecution of the employee without the employee’s consent.” Id., at 665-666. In the same vein, in Acton, 515 U. S., at 658, we relied in part on the fact that “the results of the tests are disclosed only to a limited class of school personnel who have a need to know; and they are not turned over to law enforcement authorities or used for any internal disciplinary function” in finding the searches reasonable.
“Today’s opinion speaks of a ‘elosely guarded’ class of permissible sus-picionless searches which must be justified by a ‘special need.’ But this term, as used in Skinner and Von Raab and on which the Court now relies, was used in a quite different sense than it is used by the Court today. In Skinner and Von Raab it was used to describe a basis for a search apart from the regular needs of law enforcement, Skinner, [489 U. S.], at 620; Von Raab, [489 U. SJ, at 669. The ‘special needs’ inquiry as delineated there has not required especially great ‘importan[ee],’ [520 U. SJ, at 318, unless one considers ‘the supervision of probationers,’ or the ‘operation of a government office,’ Skinner, supra, at 620, to be especially ‘important.’ Under our precedents, if there was a proper governmental purpose other than law enforcement, there was a ‘special need,’ and the Fourth Amendment then required the familiar balancing between that interest and the individual’s privacy interest.” Chandler v. Miller, 520 U. S., at 325 (Rehnquist, C. J., dissenting).
Our emphasis on this distinction should malee it clear that, contrary to the hyperbole in the dissent, we do not view these reporting requirements as “dearly bad.” See post, at 95-96, n. 3. Those requirements are simply not in issue here.
Accordingly, the police organized a meeting with the staff of the police and hospital laboratory staffs, as well as Nurse Brown, in which the police went over the concept of a chain of custody system with the MUSC staff. App. 1052-1053.
We italicize those words lest our reasoning be misunderstood. See post, at 86-88 (Kennedy, J., concurring in judgment). In none of our previous special needs cases have we upheld the collection of evidence for criminal law enforcement purposes. Our essential point is the same as Justice Kennedy’s — the extensive entanglement of law enforcement cannot be justified by reference to legitimate needs.
According to the dissent, the fact that MUSG performed tests prior to the development of Policy M-7 should immunize any subsequent testing policy despite the presence of a law enforcement purpose and extensive law enforcement involvement. See post, at 98-100. To say that any therapeutic purpose did not disappear is simply to miss the point. What matters is that under the new policy developed by the solicitor’s office and MUSC, law enforcement involvement was the means by which that therapeutic purpose was to be met. Policy M-7 was, at its core, predicated on the use of law enforcement. The extensive involvement of law enforcement and the threat of prosecution were, as respondents admitted, essential to the program's success.
Accordingly, this ease differs from New York v. Burger, 482 U. S. 691 (1987), in which the Court upheld a scheme in which police officers were used to carry out administrative inspections of vehicle dismantling businesses. That case involved an industry in which the expectation of privacy in commercial premises was “particularly attenuated” given the extent to which the industry in question was closely regulated. Id., at 700. More important for our purposes, the Court relied on the “plain administrative purposes” of the scheme to reject the contention that the statute was in fact “designed to gather evidence to enable convictions under the penal laws — .” Id., at 715. The discovery of evidence of other violations would have been merely incidental to the purposes of the administrative search. In contrast, in this case, the policy was specifically designed to gather evidence of violations of penal laws.
This case also differs from the handful of seizure cases in which we have applied a balancing test to determine Fourth Amendment reasonableness. See, e. g., Michigan Dept. of State Police v. Sitz, 496 U. S. 444,455 (1990); United States v. Martinez-Fuerte, 428 U. S. 543 (1976). First, those cases involved roadblock seizures, rather than “the intrusive search of the body or the home.” See Indianapolis v. Edmond, 531 U.S., at 54-55 (Rehnquist, C. J., dissenting); Martinez-Fuerte, 428 U. S., at 561 (“[W]e deal *84neither with searches nor with the sanctity of private dwellings, ordinarily afforded the most stringent Fourth Amendment protection”). Second, the Court explicitly distinguished the cases dealing with checkpoints from those dealing with “special needs.” Sitz, 496 U. S., at 450.
Thus, under respondents’ approach, any search to generate evidence for use by the police in enforcing general criminal laws would be justified by reference to the broad social benefits that those laws might bring about (or, put another way, the social harms that they might prevent).
It is especially difficult to argue that the program here was designed simply to save lives. Amici claim a near consensus in the medical community that programs of the sort at issue, by discouraging women who use drugs from seeking prenatal care, harm, rather than advance, the cause of prenatal health. See Brief for American Medical Association as Amicus Curiae 6-22; Brief for American Public Health Association et al. as Amici Curiae 17-21; Brief for NARAL Foundation et al. as Amici Curiae 18-19.
In fact, some MUSC staff made this distinction themselves. See PI. Exh. No. 14, Hulsey, 11-17-89, Coke Committee, 1-2 (“The use of medically indicated tests for substance abuse, obtained in conventional manners, must be distinguished from mandatory screening and collection of evidence using such methods as chain of custody, etc. . . . The question is raised as to whether pediatricians should function as law enforcement officials. While the reporting of criminal activity to appropriate authorities may be required and/or ethically just, the active pursuit of evidence to be used against individuals presenting for medical care may not be proper”).
The dissent, however, mischaracterizes our opinion as holding that “material which a person voluntarily entrusts to someone else cannot be given by that person to the police, and used for whatever evidence it may contain.” Post, at 95. But, as we have noted elsewhere, given the posture of the case, we must assume for purposes of decision that the patients did not consent to the searches, and we leave the question of consent for the Court of Appeals to determine. See n. 11, supra.
The dissent further argues that our holding “leaves law enforcement officials entirely in the dark as to when they can use incriminating evidence obtained from ‘trusted’ sources.” See post, at 95. With all due respect, we disagree. We do not address a case in which doctors independently complied with reporting requirements. Rather, as we point out above, in this case, medical personnel used the criteria set out in n. 4, supra, to collect evidence for law enforcement purposes, and law enforcement officers were extensively involved in the initiation, design, and implementation of the program. In such circumstances, the Fourth Amendment’s general prohibition against nonconsensual, warrantless, and suspieionless searches applies in the absence of consent. We decline to accept the dissent’s invitation to make a foray into dicta and address other situations not before us.