Judges LEVAL and LYNCH join the opinion and concur in separate opinions.
JOHN M. WALKER, JR., Chief Judge.Jason B. Nicholas, John Lewis, Philip Rabenbauer, Frank Solimine, Robert Pa-cini, Chester Flanders, Bennie Bates, Lymond Stephenson, Luis Mejia, Cecil Barrow, and Dominic DeRuggiero (collectively, “plaintiffs”) appeal from a judgment of the United States District Court for the Southern District of New York (Kevin T. Duffy, Judge) granting defendants’ motion to dismiss. Nicholas v. Goord, No. 01 Civ. 7891, 2004 WL 1432533 (S.D.N.Y. June 24, 2004). Plaintiffs, who filed suit under 42 U.S.C. § 1983, challenge the constitutionality under the Fourth Amendment of New York’s DNA statute, which requires certain classes of convicted felons to provide DNA samples to be maintained in a state database.
We affirm the district court’s dismissal of plaintiffs’ complaint, but rely on different reasoning. We hold that the constitutionality of New York’s DNA statute is properly analyzed under the Fourth Amendment’s “special needs” test; under that test, we find the statute constitutional.
BACKGROUND
1. Facts
Plaintiffs are felons convicted in New York who, at the time they brought this lawsuit, were incarcerated. They challenge the 1999 version of New York’s DNA statute, which requires certain classes of convicted felons to provide DNA samples to be maintained in a state database. N.Y. Exec. Law § 995 et seq. (McKinney 1999).1 New York’s law is similar to the numerous DNA-indexing statutes that have been established at both the state and federal levels; it (1) mandates the extraction of DNA samples from certain classes of convicted felons, id. § 995-c(3);2 (2) provides for DNA information obtained from those samples to be maintained in an index, or database, id.; (3) specifies that DNA samples will be analyzed only for markers “having value for law enforcement identification purposes,” *656id. § 995 — c(5);3 (4) allows for release of DNA records only in limited circumstances, id. § 995-c(6);4 (5) penalizes the unauthorized disclosure or use of DNA records, id. § 995 — f; and (6) requires that an individual’s DNA records be expunged if his conviction is reversed or if he is pardoned, id. § 995 — c(9). All nine plaintiffs have provided blood samples for purposes of the DNA index.5
Suing under 42 U.S.C. § 1988, plaintiffs claim that New York’s statute violates the Fourth Amendment, which prohibits unreasonable searches and seizures. See U.S. Const, amend. IV. They seek to have their DNA records expunged from New York’s database as well as money damages.6 In addition to defendantsappellees Goord and Lapp (“State defendants”), plaintiffs named as defendants Medilabs, Inc., and its employee Jessica Walsh, who conducted DNA sampling for the state.7
II. Proceedings Below
On February 6, 2003, Magistrate Judge Gabriel W. Gorenstein issued a report recommending that the case be dismissed. Nicholas v. Goord, No. 01 Civ. 7891, 2003 WL 256774 (S.D.N.Y. Feb. 6, 2003) (“Report-Recommendation”). He first concluded that DNA sampling under the statute constituted a “search and seizure implicating the Fourth Amendment.” Id. at *3. After extensively analyzing the relevant case law, the magistrate judge found that New York’s DNA statute was subject to the “special needs” test first articulated by Justice Blackmun in his concurrence in New Jersey v. T.L.O., 469 U.S. 325, 351, 105 S.Ct. 733, 83 L.Ed.2d 720 (1985) (Blackmun, J., concurring), and applied by this court in analyzing Con*657necticut’s DNA statute, see Roe v. Marcotte, 193 F.3d 72, 79-82 (2d Cir.1999). In Magistrate Judge Gorenstein’s view, recent Supreme Court cases “require[d] that DNA indexing statutes ... be analyzed solely in accordance with the ‘special needs’ doctrine.” Report-Recommendation, 2003 WL 256774, at *11.
Following that doctrine, the magistrate judge conducted a two-part inquiry. He first asked whether New York’s law served a “ ‘special need, beyond the normal need for law enforcement.’ ” Id. In doing so, however, he declined to rely on the special need that we had identified in Marcotte, in part because that case preceded two significant intervening Supreme Court decisions concerning the special-needs test, see id. at *14, and in part because he was unconvinced that New York’s DNA statute was meant to deter recidivism, the special need relied upon in Marcotte, see id. at *12. Ultimately, the magistrate judge concluded that the primary purpose of New York’s DNA statute was “to maintain information available to solve future crimes,” and deemed that purpose a special need. Id. at *13. The magistrate judge then applied a balancing test and found that the interests of the state in maintaining a database to aid in crime investigation outweighed the minimal intrusion on plaintiffs’ privacy interests. The magistrate judge emphasized plaintiffs’ greatly reduced expectation of privacy as prisoners, id. at *16 — *17, and the “blanket approach” of the statute, which reduced the possibility of arbitrary conduct by the state, id. at *18.
The district court reached the same conclusion by a different route. The district court first expressed skepticism as to whether the Fourth Amendment even applied, noting that it was “not necessarily convinced that the Magistrate Judge was correct to so quickly dismiss the question,” but ultimately decided, in the absence of any argument from the state, to assume that the Fourth Amendment did apply. Nicholas, 2004 WL 1432533, at *2. Rather than conducting the special-needs inquiry, however, the district court found that under the Supreme Court’s decision in United States v. Knights, 534 U.S. 112, 122 S.Ct. 587, 151 L.Ed.2d 497 (2001), a general balancing test was more appropriate. See Nicholas, 2004 WL 1432533, at *3. The district court therefore explicitly declined to follow Marcotte or the magistrate judge’s recommendation, id. at *3 n. 7, and even suggested that the statute might not survive under the special-needs test, id. at *4 (arguing that “collecting DNA is part and parcel” of the state’s general interest in law enforcement, which would not qualify as a special need).
Instead of engaging in a special-needs inquiry, therefore, the district court proceeded directly to consider the statute under the traditional Fourth Amendment balancing test. After concluding that, in light of the totality of the circumstances, the state’s significant interest in “having information readily available to aid criminal investigations” outweighed plaintiffs’ minimal interest in not having to submit their DNA to indexing, the district court dismissed the complaint. Id. at *5-*6.
This appeal followed.
DISCUSSION
We review de novo a district court’s grant of a motion to dismiss for failure to state a claim under Federal Rule of Civil Procedure 12(b)(6).8 See, e.g., W. Mohegan Tribe & Nation v. Orange Coun*658ty, 395 F.3d 18, 20 (2d Cir.2004) (per cu-riam). We accept as true the allegations in the complaint and draw all reasonable inferences in plaintiffs’ favor. Id.
I. The Applicability of the Fourth Amendment
As a preliminary matter, we reject the district court’s sua sponte suggestion that the Fourth Amendment might not apply to New York’s DNA statute because plaintiffs may not have a reasonable expectation of privacy in their DNA. See Nicholas, 2004 WL 1432533, at *2; see generally Oliver v. United States, 466 U.S. 170, 177, 104 S.Ct. 1735, 80 L.Ed.2d 214 (1984) (“The Amendment [protects] only those expectation^] that society is prepared to recognize as reasonable.” (internal quotation marks omitted)). The state did not dispute the applicability of the Fourth Amendment below, nor does it on appeal. Our sister circuits have consistently found, as we presumed in Marcotte, 193 F.3d at 77, that prisoner DNA extraction is subject to the Fourth Amendment. See United States v. Sczubelek, 402 F.3d 175, 182 (3d Cir.2005); Padgett v. Donald, 401 F.3d 1273, 1277 (11th Cir.2005); United States v. Kincade, 379 F.3d 813, 821 n. 15 (9th Cir.2004) (en banc), cert. denied, - U.S. -, 125 S.Ct. 1638, 161 L.Ed.2d 483 (2005); Green v. Berge, 354 F.3d 675, 676 (7th Cir.2004); Groceman v. DOJ, 354 F.3d 411, 413 (5th Cir.2004); Boling v. Romer, 101 F.3d 1336, 1340 (10th Cir. 1997); Jones v. Murray, 962 F.2d 302, 306 (4th Cir.1992).
Moreover, while we agree with the district court that fingerprinting and DNA indexing serve similar purposes, see Nicholas, 2004 WL 1432533, at *2 n. 4; see also infra Part III.B, and while the Supreme Court has suggested that the former may not fall within the Fourth Amendment’s scope, see United States v. Dionisio, 410 U.S. 1, 14-15, 93 S.Ct. 764, 35 L.Ed.2d 67 (1973); but cf. id. at 39, 93 S.Ct. 764 (Marshall, J., dissenting), the Court has also recognized a distinction between non-intrusive means of obtaining physical evidence (such as fingerprinting) and more invasive measures (such as drawing blood), id. at 14-15, 93 S.Ct. 764. The latter are, under Skinner, deemed Fourth Amendment searches. See 489 U.S. at 616, 109 S.Ct. 1402. The distinction between the physical intrusion required to take a fingerprint and the intrusion required to draw a blood sample is thus constitutionally significant.
Finally, to the extent that the district court relied on Second Circuit cases holding that prisoners have a reduced expectation of privacy, see, e.g., Willis v. Artuz, 301 F.3d 65, 69 (2d Cir.2002) (prisoners lack expectation of privacy in prison cell), we note that prisoners retain a right to bodily privacy, even if that right is limited by institutional and security concerns, see Marcotte, 193 F.3d at 78; Covino v. Patrissi, 967 F.2d 73, 78 (2d Cir. 1992). Drawing blood from inmates thus effects a constitutionally cognizable intrusion on prisoners’ expectation of bodily privacy, though that expectation may be diminished. Accordingly, we find that the extraction and analysis of plaintiffs’ blood for DNA-indexing purposes constituted a search implicating the Fourth Amendment.
From this point, our analysis proceeds in two parts: We first decide which Fourth Amendment test to apply to New York’s DNA statute, and then we analyze the statute’s constitutionality under that test.
II. Special-Needs Test
A. Special-Needs or General Balancing Test?
To date, both state and federal DNA-indexing statutes have withstood Fourth *659Amendment challenges.9 See Sczubelek, 402 F.3d at 184; Kincade, 379 F.3d at 830-31 & n. 25. Courts remain divided, however, as to the appropriate test to apply. The Second, Seventh, and Tenth Circuits have applied the special-needs test. See Marcotte, 193 F.3d at 78-79; Green, 354 F.3d at 677-78; United States v. Kimler, 335 F.3d 1132, 1146 (10th Cir.2003).10 The Third, Fourth, Fifth, Ninth, and Eleventh Circuits have applied a general balancing test, see Sczubelek, 402 F.3d at 184; Jones, 962 F.2d at 307; Groceman, 354 F.3d at 413; Kincade, 379 F.3d at 832; Padgett, 401 F.3d at 1280, although the Third and Ninth Circuit decisions prompted impassioned dissents, see Sczubelek, 402 F.3d at 189-204 (McKee, J., dissenting); Kincade, 379 F.3d at 842-71 (Reinhardt, J., dissenting); id. at 871-75 (Kozinski, J., dissenting); id. at 875-76 (Hawkins, J., dissenting).11
In Marcotte, we applied the special-needs test to the DNA-indexing requirement under Connecticut’s sex-offender statute, which is similar to New York’s statute.12 See Marcotte, 193 F.3d at 75 (describing Connecticut statute). Plaintiffs, as might be expected, argue that Marcotte should be followed and the more stringent special-needs test applied. Our analysis of New York’s statute must, however, take into account not only Marcotte, but also several significant intervening Supreme Court cases. Particularly relevant are the Supreme Court’s decisions in City of Indianapolis v. Edmond, 531 U.S. 32, 121 S.Ct. 447, 148 L.Ed.2d 333 (2000), Ferguson v. City of Charleston, 532 U.S. 67, 121 S.Ct. 1281, 149 L.Ed.2d 205 (2001), and Illinois v. Lidster, 540 U.S. 419, 124 S.Ct. 885, 157 L.Ed.2d 843 (2004), which clarify the circumstances in which the special-needs exception applies and how courts should apply it. Of course, we are bound by our own precedent “unless and until its rationale is overruled, implicitly or expressly, by the Supreme Court or this court en banc.” BankBoston, N.A. v. Sokolowski, 205 F.3d 532, 534-35 (2d Cir.2000) (per curiam). In light of these subsequent *660Supreme Court decisions, however, we are free to revisit Marcotte’s reasoning, and, in light of the judicial views reflected in the circuit split, we think it prudent to do so.
B. The Evolution of the Special-Needs Exception
The Fourth Amendment prohibits unreasonable searches and seizures. In the criminal-law context, a warrant and probable cause are usually required. See Mincey v. Arizona, 437 U.S. 385, 390, 98 S.Ct. 2408, 57 L.Ed.2d 290 (1978). Warrantless searches must generally fit within “a few specifically established and well-delineated exceptions,” id. (internal quotation marks omitted), such as the warrant-less search pursuant to a lawful arrest, see Chimel v. California, 395 U.S. 752, 762-63, 89 S.Ct. 2034, 23 L.Ed.2d 685 (1969). And warrantless searches must still generally be based upon probable cause, see T.L.O., 469 U.S. at 340, 105 S.Ct. 733, though the Court has recognized that probable cause, which is “peculiarly related to criminal investigations,” Bd. of Educ. of Indep. Sch. Dist. No. 92 v. Earls, 536 U.S. 822, 828, 122 S.Ct. 2559, 153 L.Ed.2d 735 (2002) (internal quotation marks omitted), is not an “irreducible requirement of a valid search,” T.L.O., 469 U.S. at 340, 105 S.Ct. 733. However, and crucially for our purposes, where neither warrant nor probable cause is required, searches must usually be based upon some quantum of individualized suspicion; suspicionless searches are constitutional “only [in] limited circumstances.” Edmond, 531 U.S. at 37, 121 S.Ct. 447.
The special-needs exception, which developed against this backdrop of Fourth Amendment requirements, was first enunciated by Justice Blackmun in his T.L.O. concurrence, in which he clarified that exceptions to the usual warrant and probable-cause requirements were appropriate only where “special needs, beyond the normal need for law enforcement, make the warrant and probable-cause requirement impracticable.” T.L.O., 469 U.S. at 351, 105 S.Ct. 733 (Blackmun, J., concurring). As originally formulated, the special-needs exception was thus aimed at evaluating the constitutionality of warrantless searches, but did not address the appropriate standard for evaluating suspicionless searches. See also Ferguson, 532 U.S. at 74 n. 7, 121 S.Ct. 1281. Indeed, T.L.O. itself involved a warrantless search based on individualized suspicion. See T.L.O., 469 U.S. at 342 n. 8, 105 S.Ct. 733 (“Because the search ... was based upon an individualized suspicion ... we need not consider the circumstances that might justify school authorities in conducting searches unsupported by individualized suspicion.”).
Warrantless searches that serve a special need and are based on individualized suspicion have been upheld by the Court several times. See, e.g., O’Connor v. Ortega, 480 U.S. 709, 725, 107 S.Ct. 1492, 94 L.Ed.2d 714 (1987) (warrantless search of employee workspace based on reasonable suspicion of employee misconduct). As the government implicitly concedes, however, the New York statute establishes a suspi-cionless-search regime, not just a warrant-less one. Cf. United States v. Lifshitz, 369 F.3d 173, 188 (2d Cir.2004) (“In Marcotte, we explicitly distinguished searches based solely on an individual’s status as a convicted sex offender from searches based upon at least some level of individualized suspicion.”). Relevant to this appeal, therefore, are those cases concerning sus-picionless searches.
Such searches, which have historically been treated as a “closely guarded category,” Chandler v. Miller, 520 U.S. 305, 309, 117 S.Ct. 1295, 137 L.Ed.2d 513 (1997), have been upheld only in limited circum*661stances, including searches conducted at the border,13 in prisons,14 and at airports and entrances to government buildings;15 administrative or regulatory searches, particularly of closely regulated businesses;16 student and employee drug tests;17 information-seeking checkpoints;18 and searches of probationers’ residences.19 See also Kincade, 379 F.3d at 822-23 (describing categories of suspicionless searches).
What unifies these cases, despite their varied contexts, is that in each instance, the Court found that, the suspicionless-search regime at issue served some special need distinct from normal law-enforcement needs. In Martinez-Fuerte, for example, the Court “emphasized the difficulty of effectively containing illegal immigration at the border itself.” Edmond, 531 U.S. at 38, 121 S.Ct. 447 (construing Martinez-Fuerte). In Bell v. Wolfish, the Court’s concern was the “significant and legitimate sécurity interests of the [prison] institutions.” 441 U.S. at 559-60, 99 S.Ct. 1861. In Von Raab, Skinner, Vemonia, and Earls — all involving drug tests — the Court found that the purpose of the regime “was one divorced from the State’s general interest in law enforcement.” Ferguson, 532 U.S. at 79, 121 S.Ct. 1281 (construing those cases); see also id. at 80 n. 16, 121 S.Ct. 1281 (noting that in each of the four cases, results of the drug tests could not be used in criminal prosecution).
Thus, although the special-needs exception was originally formulated in the context of warrantless searches, the evolution of the Court’s Fourth Amendment jurisprudence suggests that the doctrine has increasingly become the test employed by the Court in suspicionless search cases. See also Chandler, 520 U.S. at 313, 117 S.Ct. 1295 (“[Particularized exceptions to the main rule [requiring individualized suspicion] are sometimes warranted based on special needs, beyond the normal need for law enforcement.” (internal quotation marks omitted)). Indeed, in two recent *662cases, the relationship between the special-needs exception and suspicionless-search regimes has become explicit.
In 2001, the court decided Edmond, which concerned an Indianapolis checkpoint program “whose primary purpose [was] the discovery and interdiction of illegal narcotics.” 531 U.S. at 34, 121 S.Ct. 447. After noting that it had “never approved a checkpoint program whose primary purpose was to detect evidence of ordinary criminal wrongdoing,” id. at 41, 121 S.Ct. 447, the Court held the program unconstitutional, emphasizing that it had recognized “only limited exceptions to the general rule that a seizure must be accompanied by some measure of individualized suspicion,” id. Of particular importance was Edmond’s characterization of the Court’s own special-needs jurisprudence: “[W]e have upheld certain regimes of sus-picionless searches where the program was designed to serve special needs, beyond the normal need for law enforcement,” as well as limited searches for “certain administrative purposes.” Id. at 37, 121 S.Ct. 447 (internal quotation marks omitted) (emphasis added). And Chief Justice Rehnquist in dissent seemed to assume that suspicionless searches must satisfy the special-needs inquiry. See id. at 54, 121 S.Ct. 447 (Rehnquist, C.J., dissenting) (“The ‘special needs’ doctrine, which has been used to uphold suspicion-less 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.”).
The following year the Court decided Ferguson, in which the petitioners challenged a hospital program that tested their urine for cocaine use. See 532 U.S. at 71-73, 121 S.Ct. 1281. The Court found that the “immediate objective of the searches was to generate evidence for law enforcement purposes,” id. at 83, 121 S.Ct. 1281, and deemed the program unconstitutional under the special-needs test, id. at 85, 121 S.Ct. 1281. In so holding, the Court again indicated that the special-needs test applied to searches conducted in “the absence of a warrant or individualized suspicion.” 532 U.S. at 79, 121 S.Ct. 1281.
Edmond and Ferguson are notable for two reasons. First, they indicate that searches conducted in the absence of individualized suspicion are subject to the special-needs test. While the special-needs exception was originally developed in relation to the Fourth Amendment’s warrant requirement, cases like Edmond and Ferguson have increased the doctrine’s importance in a subcategory of warrantless searches — suspicionless searches. See N.G. v. Connecticut, 382 F.3d 225, 236-37 (2d Cir.2004) (noting that when it comes to “searches undertaken pursuant to a general scheme without individualized suspicion,” the Supreme Court has applied the special-needs test and has held that “a primary purpose to advance the general interest in crime control will not suffice” (internal quotation marks and citations omitted)). In this regard, we disagree with our colleagues on the Ninth Circuit, who observed in Kincade that “[t]he Court has long understood special needs analysis to be triggered not by a complete absence of suspicion, but by a departure from the Fourth Amendment’s warrant-and-probable cause requirements.” Kincade, 379 F.3d at 829. As explained above, while T.L.O. and other early special-needs cases employed the exception in the context of warrantless searches, the Court’s recent Fourth Amendment jurisprudence has increasingly associated the special-needs test with suspicionless-search regimes.
Second, Edmond and Ferguson clarify what may qualify as a special need. Edmond asserts that a program serving a *663“general interest in crime control” will not suffice. 531 U.S. at 44, 121 S.Ct. 447. Edmond rejected the state’s argument that the checkpoint program served a non-law-enforcement need because it was broadly aimed at society’s drug problem: “If we were to rest the case at this high level of generality, there would be little check on the ability of the authorities to construct roadblocks for almost any conceivable law enforcement purpose.” Id. at 42, 121 S.Ct. 447. Ferguson reiterated Edmond’s holding, finding that a search whose “immediate objective ... was to generate evidence for law enforcement purposes” was unconstitutional, even if its “ultimate goal” was to stop substance abuse by pregnant women. 532 U.S. at 82-83, 121 S.Ct. 1281. We thus read Edmond and Ferguson to call for the application of the special-needs test in cases involving suspicionless searches, and to require that such searches serve as their immediate purpose an objective distinct from the ordinary evidence gathering associated with crime investigation. See also Chandler, 520 U.S. at 314, 117 S.Ct. 1295 (defining special needs as those “other than crime detection”).
Our understanding of the special-needs doctrine and our reading of Edmond and Ferguson must, however, also take into account the Court’s more recent decision in Illinois v. Lidster. That case concerned a highway checkpoint set up by police one week after a hit-and-run accident “at about the same time of night and at about the same place” as the accident; the checkpoint was “designed to obtain more information about the accident from the motoring public.” 540 U.S. at 422,124 S.Ct. 885. In upholding the checkpoint program, the Court noted that it “differfed] significantly from that in Edmond,” emphasizing in particular that the checkpoint’s “primary law enforcement purpose was not to determine whether a vehicle’s occupants were committing a crime, but to ask vehicle occupants, as members of the public, for their help in providing information about a crime in all likelihood committed by others.” Id. at 423,124 S.Ct. 885. The Court distinguished between an “information-seeking kind of stop” like the one at issue in Lidster, and the traffic stop at issue in Edmond, which served the state’s “general interest in crime control.” Id. at 424, 124 S.Ct. 885 (internal quotation marks omitted). And, in a seeming effort to expand the boundaries of the special-needs exception, the Court expressly observed that Edmond’s prohibition on searches conducted pursuant to a “general interest in crime control” did “not refer to every law enforcement objective,” but rather only to normal law-enforcement objectives.20 Id. (internal quotation marks omitted). Lid-ster explained that not all law-enforcement concerns would be deemed to fall outside of the special-needs exception; rather, some “special law enforcement concerns will sometimes justify [checkpoint seizures] without individualized suspicion.” Id.21
*664C. Analysis
With Edmond, Ferguson, and Lidster in mind, we now consider plaintiffs’ contention that the district court erred in applying the traditional Fourth Amendment balancing test, rather than the special-needs test. Plaintiffs urge us to follow the methodology employed in Marcotte and to apply the more stringent special-needs test.22 Certainly, the Court’s emphasis in its recent cases on applying the special-needs test to suspicionless searches strongly suggests, if it does not require, that we should continue to apply the special-needs test to DNA-indexing statutes as we did in Marcotte.
Defendants, however, maintain that a general Fourth Amendment balancing inquiry is more appropriate and argue that we should affirm the district court’s approach. Defendants offer two primary reasons for departing from Marcotte, which we now consider.23
1. United States v. Knights
Defendants first contend that the Supreme Court’s decision in United States v. Knights, 534 U.S. 112, 122 S.Ct. 587, 151 L.Ed.2d 497 (2001), justifies application of a traditional balancing test. That case, on which several courts, including the district court, have relied in dispensing with the special-needs test, see Nicholas, 2004 WL 1432533, at *3; Sczubelek, 402 F.3d at 186; Padgett, 401 F.3d at 1279-80; Kincade, 379 F.3d at 830; Groceman, 354 F.3d at 413, concerned a probationer whose house and residence the police searched. The police had long suspected Knights of being involved in various acts of vandalism. Knights, 534 U.S. at 114, 122 S.Ct. 587. The Court observed that the search, which was conducted pursuant to a search condition of Knights’s probation order,24 was *665“supported by reasonable suspicion.” Id. In upholding the search, the Court explained that searches of probationers conducted pursuant to probation conditions did not need to serve a special need, such as a probationary purpose, to be deemed constitutional. Id. at 117-18, 122 S.Ct. 587. The Court found that, in light of the reasonable suspicion supporting the search, and in light of Knights’s “significantly diminished ... expectation of privacy,” id. at 120, 122 S.Ct. 587, the search satisfied the Fourth Amendment’s reasonableness requirement.
Courts that have relied upon Knights as justifying the application of a general balancing test to DNA-indexing statutes have emphasized Knights’s status as a probationer and his knowledge of the probation search condition, which reduced his expectation of privacy. See, e.g., Sczubelek, 402 F.3d at 183 (noting that Court had viewed Knights’s probation search condition as the “ ‘salient circumstance’ ” (quoting Knights, 534 U.S. at 118, 122 S.Ct. 587)); Padgett, 401 F.3d at 1279 (“Key to the Court’s ruling was Knights’ status as probationer.”); Kincade, 379 F.3d at 827-28. Because the Court emphasized Knights’s status as a probationer subject to a probation search condition, and because the Court applied a general balancing inquiry, our sister circuits have assumed that those two facts are causally related. They have thus interpreted the Court’s willingness to employ a general balancing test, rather than the special-needs test, as an indication that a general balancing test is appropriate wherever the person searched has a reduced expectation of privacy, regardless of whether the search is supported by individualized suspicion. See, e.g., id. at 832.
We are unwilling to leap to that conclusion. The Court’s decision to employ a traditional balancing test in Knights must be viewed in context. In particular, we think it telling that the Court emphasized, from the very first paragraph of its opinion, that the search of Knights’s apartment was “supported by reasonable suspicion.” Knights, 534 U.S. at 114, 122 S.Ct. 587; see also id. at 121-22,122 S.Ct. 587 (noting repeatedly that the search was supported by reasonable suspicion). And this court has previously construed Knights as limited to situations involving some quantum of individualized suspicion. See Lifshitz, 369 F.3d at 180-81 (“In Knights ... [the Court] decided only that the particular search at issue met the requirements of the Fourth Amendment, because it was based on reasonable suspicion.”). Indeed, the Court expressly noted that it was “not addressing] the constitutionality of a sus-picionless search because the search in this case was supported by reasonable suspicion.” Knights, 534 U.S. at 120 n. 6, 122 S.Ct. 587. In light of the Court’s emphasis on the existence of reasonable suspicion in Knights, its decision to employ a general balancing test in that case was arguably due as much to the existence of individualized suspicion as it was to Knights’s reduced expectation of privacy under the search condition. As Judge Reinhardt explained in his Kincade dissent, in Knights,
The Court distinguished the “special needs” line of cases, but it did so cautiously, explaining that its departure from that framework was justified only by the combination of all of the circumstances present. Those circumstances included the reduced expectation of privacy held by Knights on account of the conditions of his probation. [They] also included, as the Court emphasized repeatedly, the fact that the search was supported by reasonable suspicion[.]
*666Kincade, 379 F.3d at 861 (Reinhardt, J., dissenting); see also Sczubelek, 402 F.3d at 196 (McKee, J., dissenting) (“[T]here was no real issue [in Knights ] about whether the search was justified by a reasonable suspicion .... Rather, the issue was whether the warrantless search of Knights’ private residence. [was constitutional].”).
We thus reject defendants’ argument that Knights justifies applying the traditional Fourth Amendment balancing test to New York’s DNA statute. In light of the Court’s emphasis in its recent Fourth Amendment cases on applying the special-needs test to suspicionless-search regimes, see, e.g., Chandler, 520 U.S. at 313-14, 117 S.Ct. 1295; Edmond, 531 U.S. at 37, 121 S.Ct. 447, as well as the Court’s focus in Knights on the existence of reasonable suspicion in that case, we decline to construe Knights as permitting us to apply a general balancing test to suspicionless searches. The Supreme Court has never applied a general balancing test to a suspi-cionless-search regime. See Kincade, 379 F.3d at 862 (Reinhardt, J., dissenting). Until the Supreme Court expressly adopts such an approach, the more prudent route, and the route more consonant with the values underlying the Fourth Amendment, see Vernonia, 515 U.S. at 667, 115 S.Ct. 2386 (O’Connor, J., dissenting) (“For most of our constitutional history, mass, suspi-cionless searches have been generally considered per se unreasonable within the meaning of the Fourth Amendment.”), is to construe Knights as dispensing with the special-needs test not solely because of the probation search condition, but also because of the existence of individualized suspicion. See Kincade, 379 F.3d at 863 (Reinhardt, J., dissenting).
2. Reduced Expectation of Privacy
Defendants also contend that the searches conducted in Ferguson and Edmond are distinguishable from the search at issue in this case, and therefore those cases do not apply. Specifically, ‘defendants argue that neither Edmond nor Ferguson involved a “class of individuals [that] has a diminished expectation of privacy,” whereas this case involves prison inmates, who have a “substantially diminished expectation of privacy in their identifying information.”25 See also Kincade, 379 F.3d at 832; Nicholas, 2004 WL 1432533, at *3. In other words, defendants and those courts that have applied Knights argue that while suspicionless searches of the general public may require scrutiny under the special-needs test, suspicionless searches of individuals with a reduced expectation' of privacy are subject only to the more lenient balancing test.
The- problem with this argument is that neither Ferguson nor Edmond rested upon the plaintiffs’ undiminished expectation of privacy. Rather, the key to each case was the program’s law-enforcement purpose. See Ferguson, 532 U.S. at 79, 121 SICt. 1281 (noting that the “critical difference between [earlier] drug-testing cases and this one ... lies in the nature of the ‘special need’ asserted as justification for the warrantless searches”); id. at 83-84, 121 S.Ct. 1281 (finding “critical” the fact that the “immediate objective of the searches was to generate evidence for law enforcement purposes ” (emphasis in original)); Edmond, 531 U.S. at 48, 121 S.Ct. 447 (“Because the primary purpose of the Indianapolis checkpoint program is ultimately indistinguishable from the general interest in crime control, the checkpoints violate the Fourth Amendment.”). Moreover, as plaintiffs point out, almost every *667special-needs case considered by the Supreme Court has involved individuals with a diminished expectation of privacy. See, e.g., Vernonia, 515 U.S. at 657, 115 S.Ct. 2386; Earls, 536 U.S. at 830, 122 S.Ct. 2559. Indeed, this court has previously held that a “diminished expectation of privacy” is a principal criterion of special-needs cases. Lifshitz, 369 F.3d at 186 (“[T]hose subject to the search must enjoy a diminished expectation of privacy, partly occasioned by the special nature of their situation, and partly derived from the fact that they are notified in advance of the search policy.”).
We therefore cannot agree with defendants’ contention that a reduced expectation of privacy allows courts to dispense with the special-needs test in cases involving suspicionless-search regimes. Indeed, we view such logic with some concern, in light of the wide swath of the general public who at one point or another has had a reduced expectation of privacy. See, e.g., Lidster, 540 U.S. at 424, 124 S.Ct. 885 (motorists); Earls, 536 U.S. at 831-32, 122 S.Ct. 2559 (student athletes and students participating in extracurricular activities); Vernonia, 515 U.S. at 656, 115 S.Ct. 2386 (public-school students); Von Raab, 489 U.S. at 672, 109 S.Ct. 1384 (government employees involved in drug interdiction); Skinner, 489 U.S. at 627, 109 S.Ct. 1402 (employees participating “in an industry that is regulated pervasively to ensure safety”). Were we to apply the general balancing test to New York’s statute simply because the individuals searched had a diminished expectation of privacy, we would be approving the application of a considerably more lenient standard of review to suspicionless-search regimes that have heretofore been subject to a more searching inquiry. We decline so to relax our review of such regimes, which have historically been regarded as a “closely guarded category.” Chandler, 520 U.S. at 309, 117 S.Ct. 1295; see also Edmond, 531 U.S. at 37, 121 S.Ct. 447.
We therefore reaffirm the approach we took in Marcotte and conclude that plaintiffs’ Fourth Amendment challenge to New York’s DNA-indexing statute is properly analyzed under the special-needs test. Although courts have unanimously upheld DNA-indexing statutes whether they have applied the special-needs test or the general Fourth Amendment balancing test, the test applied continues to matter, especially since the reasons for adopting a particular test will inevitably have consequences in other search contexts. We therefore continue to hold suspicionless searches to the higher standard of review embodied in the special-needs inquiry.
III. Analysis Under the Special-Needs Test
A. Does New York’s Statute Serve a Special Need?
In determining whether New York’s DNA statute can be justified under the special-needs exception, we first ask what the statute’s primary purpose is, mindful that it is the statute’s immediate rather than ultimate objective that is relevant. See Ferguson, 532 U.S. at 82-83, 121 S.Ct. 1281. In Marcotte, we considered the constitutionality of Connecticut’s DNA statute, which, inter alia, required covered sex offenders to submit a blood sample for DNA indexing. See 193 F.3d at 74. We recognized that the statute was “not motivated by concerns for inmate safety and health, institutional order, or discipline” that have usually supported a special-needs exception in the prison context. Id. at 78. Nevertheless, we found that the statute did serve special needs, in that it would (1) aid law enforcement in solving past and future crimes, and (2) deter recidivism. Id. at 79.
*668Although we conclude that New York’s statute likewise serves a special need, distinguishable from ordinary law-enforcement needs, we do not think that the immediate objective of the statute is to deter recidivism, although such deterrence may be a valuable byproduct.26 We instead agree with the magistrate judge that close examination of the statute reveals that its “primary purpose is to create a DNA database to assist in solving crimes should the investigation of such crimes permit resort to DNA testing of evidence.” Report-Recommendation, 2003 WL 256774, at *12. The website of New York’s Division of Criminal Justice Services states that “[t]he primary function of the DNA Databank is to maintain DNA profiles of convicted offenders that can be used by law enforcement to identify a perpetrator of a crime when DNA evidence is retrieved from a crime scene.” Joint Appendix (“JA”) at 35.27 By contrast, the website makes no mention of deterring recidivism and, as the magistrate judge pointed out, the legislative history surrounding both the 1994 enactment of the statute and its 1999 amendment is “devoid of references to identifying human remains and discouraging recidivism.” Report-Recommendation, 2003 WL 256774, at *12.
Significant also are those provisions of the New York statute authorizing release of DNA records. The statute allows DNA records to be released only (1) to law-enforcement agencies “upon submission of a DNA record in connection with the investigation of the commission of one or more crimes or to assist in the recovery or identification of specified human remains”; or (2) for “criminal defense purposes,” where a defendant seeks access to “samples and analyses performed in connection with the case.” Id. § 995 — c(6).28 As the magistrate judge recognized, unless we think identification of human remains is the primary purpose of the statute (an unlikely prospect), the release provisions indicate that providing information to aid in investigations is the statute’s immediate objective. See Report-Recommendation, 2003 WL 256774, at *11.
We therefore ask whether a DNA-indexing statute that aims to create a DNA-identification index to assist in solving crimes serves a special need, as that term has been defined by the Court in its recent cases. There can be little doubt that New York’s statute serves a purpose related to law enforcement, but we do not think that fact automatically condemns the New York statute. In light of the distinction drawn by the Court in Lidster between “information-seeking” searches or seizures, which respond to “special law enforcement concerns,” 540 U.S. at 424, 124 S.Ct. 885, and those regimes aimed at “detecting] evidence of ordinary criminal wrongdoing,” id. at 423, 124 S.Ct. 885 (internal quotation marks omitted), we think a more nuanced approach to law-enforcement concerns is appropriate. Lidster instructs courts to examine carefully the type of law-enforcement concern served by a particular search or seizure regime. Like the magistrate judge, we find it crucial that the state, in collecting DNA samples, is not trying to “determine that a particular individual has engaged in some specific wrongdoing.” Report-Recommen*669dation, 2003 WL 256774, at *13. Although the DNA samples may eventually help law enforcement identify the perpetrator of a crime, at the time of collection, the samples “in fact provide no evidence in and of themselves of criminal wrongdoing,” and are not sought “for the investigation of a specific crime.” Id. (internal quotation marks omitted). Because the state’s purpose in conducting DNA indexing is distinct from the ordinary “crime detection” activities associated with normal law-enforcement concerns, it meets the special-needs threshold. See Green, 354 F.3d at 678 (“Although the state’s DNA testing of inmates is ultimately for a law enforcement goal, ... it is not undertaken for the investigation of a specific crime” (internal quotation marks omitted)); Kimler, 335 F.3d at 1146 (“[U]nder the special needs exception ... the desire to build a DNA database goes beyond the ordinary law enforcement need.”).
B. Special-Needs Balancing Test
Having concluded that New York’s DNA statute serves a special need, we now weigh that special need against the privacy intrusion it effects to determine whether it is reasonable within the meaning of the Fourth Amendment. See O’Connor, 480 U.S. at 725-26, 107 S.Ct. 1492; see also Lidster, 540 U.S. at 426-27, 124 S.Ct. 885. We conduct a “fact-specific balancing of the intrusion on the ... Fourth Amendment rights [of the persons searched] against the promotion of legitimate governmental interests.” Earls, 536 U.S. at 830, 122 S.Ct. 2559; see also Chandler, 520 U.S. at 314, 117 S.Ct. 1295.
There can be little doubt that New York has a strong government interest in obtaining identifying information from convicted offenders and keeping a record of such information. See, e.g., Sczubelek, 402 F.3d at 185 (“The interest in accurate criminal investigations and prosecutions is a compelling interest that the DNA Act can reasonably be said to advance.”); Kincade, 379 F.3d at 838-39 (finding that federal DNA statute serves “undeniably compelling” state interests in (1) identifying probationers who commit crimes once they are at large, and (2) deterring recidivism); Green, 354 F.3d at 679 (finding that Wisconsin’s DNA statute “serves an important state interest” in allowing law enforcement to collect “the most reliable evidence of identification”); see also Earls, 536 U.S. at 824, 122 S.Ct. 2559 (evaluating strength of government’s interest as well as “efficacy” of program in serving that interest). Nor is there any question that New York’s statute is effective in advancing that state interest.
Against these government interests, the court must weigh the intrusion on inmates, which is twofold. First, offenders are subject to a physical intrusion when they are required to provide the DNA sample, whether by blood sample or buccal cheek swab. We conclude that this physical intrusion is far outweighed by the government’s strong interests in obtaining from plaintiffs the uniquely effective identifying information that DNA provides. The Supreme Court has long maintained that the intrusion effected by taking a blood sample, while subject to the Fourth Amendment, is minimal. See Skinner, 489 U.S. at 624, 109 S.Ct. 1402. In the prison context, where inmates are routinely subject to medical procedures, including blood draws, and where their expectation of bodily privacy, while intact, is diminished, see Marcotte, 193 F.3d at 78, the intrusiveness of a blood draw is even further minimized.29
*670The second intrusion to which offenders are subject is the analysis and maintenance of their DNA information in New York’s database. This intrusion may be viewed either as a search, see Skinner, 489 U.S. at 616-17, 109 S.Ct. 1402 (chemical analysis of blood sample “to obtain physiological data” is a Fourth Amendment séarch), or as a seizure, see id. at 617 & n. 4, 109 S.Ct. 1402. Regardless, it is potentially a far greater intrusion than the initial extraction of DNA, since the state analyzes DNA for information and maintains DNA records indefinitely. It is this intrusion that has caused’the greatest concern among those of our colleagues who would strike down DNA-indexing statutes as unconstitutional. See Kincade, 379 F.3d at 867 (Reinhardt, J., dissenting) (arguing that DNA indexing “constitutes far more of an intrusion than the mere insertion of a needle,” since the samples are turned into “profiles capable of being searched time and time again throughout the course of an individual’s life”); id. at 872 (Kozinski, J., dissenting) (“[I]f we accept the legal presumption ... that once [an offender] leaves supervised release he will be just like everyone else, authorizing the extraction of his DNA now to help solve crimes later is a huge end run around the Fourth Amendment.”); see also Sczubelek, 402 F.3d at 201 (McKee, J., dissenting) (“In order to sustain the DNA search of Sczubelek, we must conclude that it is reasonable to catalogue his DNA even though he has committed no new crimes because of the possibility, however remote or theoretical, that he may one day commit another crime.”).
Although we acknowledge these concerns, we ultimately conclude that the intrusion into plaintiffs’ privacy resulting from state’s practice of analyzing and maintaining DNA records does not outweigh the government’s strong interests. Although DNA indexing has the potential to be broadly revealing, the New York statute as written does not provide for sensitive information to be analyzed or kept in its database. Rather, it provides only for the analysis of identifying markers. N.Y. Exec. Law § 995-c(3), (5). The junk DNA that is extracted has, at present, no known function, except to accurately and uniquely establish identity. Although science may someday be able to unearth much more information about us through our junk DNA, that capability does not yet exist, and, more importantly, the New York statute prohibits such analysis. Id. The law provides that DNA records “shall be confidential,” id. § 995-d(l), and criminally punishes (1) the intentional disclosure of DNA records to unauthorized individuals or entities, (2) the intentional use or receipt of DNA records for “purposes other than those authorized [by the statute],” and (3) knowingly tampering or attempting to tamper with any DNA sample or the collection container without lawful authority, id. § 995 — f; 1999 N.Y. Laws, ch. 560, § 6 (amending the statute to include anti-tampering provision).30 Although plaintiffs and amici suggest that New York’s statute could permit the state to use DNA for more harmful purposes than maintaining an identification database, those facts are not present here, and if they should arise, no doubt a different *671calculus under the special-needs analysis would result.
Given the limits imposed on the collection, analysis, and use of DNA information by the statute, we see the intrusion on privacy effected by the statute as similar to the intrusion wrought by the maintenance of fingerprint records.31 See Kincade, 379 F.3d at 836 n. 31 (noting that “everyday ‘booking’ procedures routinely require even the merely accused to provide fingerprint identification, regardless of whether investigation of the crime involves fingerprint evidence” (internal quotation marks omitted)); Green, 354 F.3d at 680 (Easterbrook, J., concurring) (“Collecting felons’ DNA, like collecting their fingerprints, handwriting exemplars, and other information that may help solve future crimes (and thus improve the deterrent force of the criminal sanction) is rationally related to the criminal conviction.”); cf. Sczubelek, 402 F.3d at 185 (“Individuals on supervised release cannot reasonably expect to keep information bearing on their physical identity from government records.”). The collection and maintenance of DNA information, while effected through relatively more intrusive procedures such as blood draws or buccal cheek swabs, in our view plays the same role as fingerprinting. Given that the state likely already has a plethora of identifying information about plaintiffs, in light of their status as convicted felons, see Reportr-Rec-ommendation, 2003 WL 256774, at *16, the additional intrusion effected by the DNA statute is insufficient to outweigh the state’s strong interest in maintaining a DNA index.32 In other words, plaintiffs’ status as convicted felons renders minimal the degree to which the New York statute intrudes on their privacy.
We therefore conclude that New York’s statute, which serves a special need beyond the normal need for law enforcement, is supported by strong government interests that outweigh the relatively minimal intrusion on plaintiffs’ expectation of privacy. Moreover, we reject plaintiffs’ argument that the state should be required to obtain a warrant before taking DNA samples. T.L.O., 469 U.S. at 351, 105 S.Ct. 733 (Blackmun, J., concurring) (noting that state must generally show that obtaining a warrant would be “impracticable” when it conducts warrantless search under the special-needs exception). Obtaining a warrant requires probable cause, see, e.g., Mincey, 437 U.S. at 390, 98 S.Ct. 2408, which obviously does not exist in the context of suspicionless searches; requiring a warrant from law enforcement would thus plainly be “impracticable.” The usual purpose of obtaining a warrant — to permit the state to engage in the normal law-enforcement function of crime investigation — is absent in the context of an information-gathering search. And the concerns that usually animate the warrant requirement — that the state will exercise its search and seizure powers arbitrarily, *672see Skinner, 489 U.S. at 622, 109 S.Ct. 1402 — are not at play in the ease of DNA-indexing statutes, which take a blanket approach and apply to all convicted offenders falling within certain categories. We further note that in applying the special-needs exception in other eases, the Supreme Court has not always required an express finding that obtaining a warrant would be impracticable. See, e.g., Lidster, 540 U.S. at 427-28, 124 S.Ct. 885; Earls, 536 U.S. at 829-37, 122 S.Ct. 2559.
Relying on the special-needs test rather than the general balancing test employed by the district court, we hold that New York’s DNA statute satisfies the Fourth Amendment. Accordingly, the district court properly granted defendants’ motion to dismiss.
CONCLUSION
For the foregoing reasons, the judgment of the district court is Affirmed.
. The statute, originally enacted in 1994, at first applied only to individuals convicted after January 1, 1996. 1994 N.Y. Laws, ch. 737, §§ 1, 3. In 1999, the statute was amended to apply to persons already convicted of certain offenses who were still serving a sentence. 1999 N.Y. Laws, ch. 560, § 9. Plaintiffs, all of whom were convicted before 1996 and were serving their sentences in 1999, became subject to the statute at that time. References to the statute throughout this opinion are entirely to the 1999 version. The parties agree that subsequent amendments are not at issue.
. The 1999 statute applied only to certain felonies (e.g., assault, homicide, rape, incest, escape, attempted murder, kidnaping, arson, burglary). See N.Y. Exec. Law § 995(7).
. DNA databases like New York's utilize "junk DNA,” which does not (as far as we know) contain genetic information. See United States v. Kincade, 379 F.3d 813, 818 (9th Cir.2004) (en banc), cert. denied, - U.S. -, 125 S.Ct. 1638, 161 L.Ed.2d 483 (2005).
. Records may only be released (1) to law-enforcement agencies for identification of specified human remains or for identification purposes in criminal investigations, (2) to a defendant or his legal representative, or (3) after personally identifiable information has been removed, to authorized entities for the purpose of maintaining a population-statistics database. N.Y. Exec. Law § 995-c(6).
. Although the statute originally required that DNA be extracted by blood sample, see 1994 N.Y. Laws, ch. 737, § 3, the statute was amended in 1999 to require only "a sample appropriate for DNA testing,” 1999 N.Y. Laws, ch. 560, § 3. The state maintains that its "current normal practice ... is to [obtain DNA by taking] [bjuccal cheek swab[s],” but conceded at oral argument that plaintiffs have all had their blood drawn. We therefore confine our analysis to the extraction of plaintiffs' DNA via blood sample. In any event, even less intrusive measures of obtaining physiological data, such as cheek swabs, can constitute a search, since "[t]he ensuing chemical analysis of the sample” may also effect an "invasion of the [searchee’s] privacy interests.” Skinner v. Ry. Labor Executives’ Ass’n, 489 U.S. 602, 616, 109 S.Ct. 1402, 103 L.Ed.2d 639 (1989).
. At the time of filing, two plaintiffs had not yet had their blood drawn; they initially sought to bar the state from doing so. At oral argument, however, the parties informed the court that all nine plaintiffs have had their blood drawn for DNA-indexing purposes. We therefore understand that all plaintiffs now seek the same remedies.
. Private parties are subject to the Fourth Amendment if they act as agents of the state. See Skinner, 489 U.S. at 614, 109 S.Ct. 1402; United States v. Bennett, 709 F.2d 803, 805 (2d Cir.1983). Under 42 U.S.C. § 1983, private parties acting under color of state law can be held liable for violations of federal constitutional rights. See Adickes v. S.H. Kress & Co., 398 U.S. 144, 152, 90 S.Ct. 1598, 26 L.Ed.2d 142 (1970); Fries v. Barnes, 618 F.2d 988, 990-91 (2d Cir.1980).
. Medilabs and Walsh submitted a motion for judgment on the pleadings, pursuant to Federal Rule of Civil Procedure 12(c), but such a motion is evaluated under the same standard as a Rule 12(b)(6) motion to dismiss. See *658Sheppard v. Beerman, 18 F.3d 147, 150 (2d Cir. 1994).
. Courts have also upheld DNA-indexing statutes in the face of other constitutional challenges. See, e.g., Doe v. Moore, 410 F.3d 1337 (11th Cir.2005) (rejecting due-process, separation-of-powers, and equal-protection challenges to DNA-indexing provisions of Florida’s sex-offender statute); Sczubelek, 402 F.3d at 187-89 (rejecting separation-of-powers challenge to federal statute); Padgett, 401 F.3d at 1280-81 (rejecting substantive due-process challenge to Georgia's DNA statute).
. Prior to Kimler, the Tenth Circuit applied a' general balancing test to Colorado’s DNA statute. See Boling, 101 F.3d at 1340. Kim-ler, however, applied the special-needs test without commenting on Boling. See 335 F.3d at 1146.
. Indeed, it is not so clear that proponents of the general balancing test prevailed in the Ninth Circuit. Kincade was decided by eleven judges sitting en banc. Five judges voted to uphold the federal DNA statute under a general balancing test, and a sixth judge voted to uphold the statute under the special-needs exception. The five dissenters would have applied the special-needs test to strike down the statute. See Kincade, 379 F.3d at 842 n. 1 (Reinhardt, J., dissenting). Thus, the special-needs test received six votes in Kincade, while the general balancing test only received five. See also Moreno v. Baca, 400 F.3d 1152, 1157 n. 2 (9th Cir.2005) ("[In Kincade ], we were unable to resolve the proper test to be applied ....”).
.Specifically, Connecticut’s statute, Conn. Gen.Stat. §. 54-102g, applies to persons convicted of certain offenses after October 1, 1994 and sentenced to incarceration, as well as those convicted before October 1, 1994 of certain offenses, but incarcerated at that time. Marcotte, 193 F.3d at 75. Connecticut's statute requires that DNA be obtained through blood sample, specifies that DNA be tested for identification purposes, and provides that the results of DNA testing be kept confidential. Id.
. See United States v. Martinez-Fuerte, 428 U.S. 543, 96 S.Ct. 3074, 49 L.Ed.2d 1116 (1976) (finding constitutional reasonably located permanent checkpoints for brief stops and questioning, for purpose of policing the border).
. See Bell v. Wolfish, 441 U.S. 520, 99 S.Ct. 1861, 60 L.Ed.2d 447 (1979) (upholding, on penological and institutional-safety grounds, suspicionless visual body-cavity searches of inmates following contact visits).
. See Chandler, 520 U.S. at 323, 117 S.Ct. 1295 ("[WJhere the risk to public safety is substantial and real, blanket suspicionless searches calibrated to the risk may rank as 'reasonable’ — for example, searches now routine at airports and at entrances to courts and other official buildings.”).
. See New York v. Burger, 482 U.S. 691, 107 S.Ct. 2636, 96 L.Ed.2d 601 (1987) (upholding administrative inspection of automobile junkyards); Camara v. Mun. Ct. of S.F., 387 U.S. 523, 87 S.Ct. 1727, 18 L.Ed.2d 930 (1967) (upholding municipal area inspections to monitor compliance with building safety codes).
. See Earls, 536 U.S. 822, 122 S.Ct. 2559 (upholding random, suspicionless drug testing of students involved in extracurricular activities); Vernonia Sch. Dist. 47J v. Acton, 515 U.S. 646, 115 S.Ct. 2386, 132 L.Ed.2d 564 (1995) (same); Nat'l Treasury Employees Union v. Von Raab, 489 U.S. 656, 109 S.Ct. 1384, 103 L.Ed.2d 685 (1989) (upholding sus-picionless drug testing of certain customs employees); Skinner, 489 U.S. 602, 109 S.Ct. 1402 (upholding testing of railroad employees involved in train accidents).
. See Lidster, 540 U.S. 419, 124 S.Ct. 885 (upholding brief stops of motorists at checkpoint where police sought information about recent hit-and-run accident).
. See Griffin v. Wisconsin, 483 U.S. 868, 107 S.Ct. 3164, 97 L.Ed.2d 709 (1987) (upholding searches of probationer's home because of state’s interest in supervising probationers).
. The dissenters also recognized “a valid and important distinction between seizing a person to determine whether she has committed a crime and seizing a person to ask whether she has any information.” Id. at 428, 124 S.Ct. 885 (Stevens, J., concurring in part and dissenting in part). They agreed that in the latter case, Edmond did not require that the seizure be deemed per se unconstitutional. Id.
. At oral argument, plaintiffs argued that Lidster is inapplicable in this case because (1) Lidster concerned a seizure, and (2) the Court has treated suspicionless checkpoint seizures differently from suspicionless searches. We reject the argument. Plaintiffs are correct insofar as the Court, in a footnote in Ferguson, observed that it had historically distinguished between "seizure cases in which we have applied a balancing test,” and cases involving searches of the body or the home, in *664which it had applied the special-needs test. 532 U.S. at 83 n. 21, 121 S.Ct. 1281. But Lidster, which postdates Ferguson, does not distinguish between searches and seizures in discussing the special-needs test. Rather, Lidster discusses the scope of the special-needs exception and Edmond without reference to any distinction between searches and seizures. Moreover, drawing a sharp distinction between search cases and seizure cases is unhelpful in this instance because the extraction and analysis of bodily fluids may constitute a seizure as well as a search. See Skinner, 489 U.S. at 616-17 & n. 4, 109 S.Ct. 1402.
.We find puzzling the Third Circuit's comment in Sczubelek that the special-needs inquiry is less rigorous than the general balancing test. See 402 F.3d at 184. The special needs exception requires the court to ask two questions. First, is the search justified by a special need beyond the ordinary need for normal law enforcement? Second, if the search does serve a special need, is the search reasonable when the government’s special need is weighed against the intrusion on the individual’s privacy interest? See N.G., 382 F.3d at 230-31; Report-Recommendation, 2003 WL 256774, at *15. A general balancing test, on the other hand, only requires the court to balance the government’s interest in conducting the search against the individual’s privacy interests. See, e.g., Kincade, 379 F.3d at 836.
. The district court also relied on Bell v. Wolfish, 441 U.S. 520, 99 S.Ct. 1861, 60 L.Ed.2d 447 (1979), for the proposition that a balancing test may be used on searches of inmates conducted on "less than probable cause." See Nicholas, 2004 WL 1432533, at *3. We reject the argument that Wolfish justifies application of a balancing test to prisoners, given that the Supreme Court itself, as well as this court, have characterized Wolfish as a special-needs case because of its focus on institutional safety concerns unique to prisons. See Skinner, 489 U.S. at 619, 109 S.Ct. 1402 (listing Wolfish as a special-needs case); N.G., 382 F.3d at 231; see also Wolfish, 441 U.S. at 559, 99 S.Ct. 1861 (noting “serious security dangers” of detention facilities).
. The probation order required that Knights, inter alia, submit his "person, property, place of residence, vehicle, personal effects, to search at any time, with or without a search *665warrant, warrant of arrest or reasonable cause.” 534 U.S. at 114, 122 S.Ct 587.
. At the time of filing, all plaintiffs were incarcerated. We discuss further below the fact that some plaintiffs have since been released.
. The Marcotte court was not alone in relying on a deterrence rationale. See Kincade, 379 F.3d at 840 (Gould, J., concurring).
. As of November 22, 2005, the website was . currently available at http://criminal justice .state.ny.us/forensic /dnafaqs.htm.
.The statute also provides for release of DNA records for research and statistical purposes, but only “after personally identifiable information has been removed.” Id. § 995-c(6)(c).
. In this regard, we note that plaintiffs have submitted materials indicating that they were *670subject to blood tests when they first entered prison. See JA at 49 ("The taking of a DNA sample involves a similar procedure to the one that was used on you when you first entered the system and a blood sample was taken from your arm by medical personnel.’’).
. These offenses, previously misdemeanors, were made class E felonies in 1999. 1999 N.Y. Laws, ch. 560, § 5.
. The analogy we draw here between fingerprinting and DNA indexing is not inconsistent with our conclusion, in Part I, supra, that the two practices are dissimilar for purposes of determining whether the Fourth Amendment is implicated. We disagreed above, in light of Dionisio and Skinner, with the district court's suggestion that the physical intrusion caused by drawing blood could be considered identical to the state’s practice of taking fingerprints. That conclusion, however, does not preclude us from finding that the state’s purpose in keeping DNA records is comparable to the state’s purpose in keeping fingerprints and photographs.
. Thus, even though some plaintiffs are no longer prisoners, and may thus claim a greater expectation of privacy than they held while incarcerated, we still find that — in light of the fact that the state regularly maintains identifying records of former inmates — the privacy intrusion remains relatively minimal.