JUSTICE SOUTER, dissenting.

I would hold that using the dog for the purposes of determining the presence of marijuana in the car's trunk was a search unauthorized as an incident of the speeding stop and unjustified on any other ground. I would accordingly affirm the judgment of the Supreme Court of Illinois, and I respectfully dissent.

In United States v. Place, 462 U.S. 696 (1983), we categorized the sniff of the narcotics-seeking dog as " sui generis" under the Fourth Amendment and held it was not a search. Id., at 707. The classification rests not only upon the limited nature of the intrusion, but on a further premise that experience has shown to be untenable, the assumption that trained sniffing dogs do not err. What we have learned about the fallibility of dogs in the years since Place was decided would itself be reason to call for reconsidering Place's decision against treating the intentional use of a trained dog as a search. The portent of this very case, however, adds insistence to the call, for an uncritical adherence to Place would render the Fourth Amendment indifferent to suspicionless and indiscriminate sweeps of cars in parking garages and pedestrians on sidewalks; if a sniff is not preceded by a seizure subject to Fourth Amendment notice, it escapes Fourth Amendment review entirely unless it is treated as a search. We should not wait for these developments to occur before rethinking Place's analysis, which invites such untoward consequences.

At the heart both of Place and the Court's opinion today is the proposition that sniffs by a trained dog are sui generis because a reaction by the dog in going alert is a response to nothing but the presence of contraband. See ibid. ("[T]he sniff discloses only the presence or absence of narcotics, a contraband item"); ante, at 409 (assuming that "a canine sniff by a well-trained narcotics-detection dog" will only reveal "`the presence or absence of narcotics, a contraband item'" (quoting Place, supra, at 707)). Hence, the argument goes, because the sniff can only reveal the presence of items devoid of any legal use, the sniff "does not implicate legitimate privacy interests" and is not to be treated as a search. Ante, at 409.

The infallible dog, however, is a creature of legal fiction. Although the Supreme Court of Illinois did not get into the sniffing averages of drug dogs, their supposed infallibility is belied by judicial opinions describing well-trained animals sniffing and alerting with less than perfect accuracy, whether owing to errors by their handlers, the limitations of the dogs themselves, or even the pervasive contamination of currency by cocaine. See, e.g., United States v. Kennedy, 131 F. 3d 1371, 1378 (CA10 1997) (describing a dog that had a 71% accuracy rate); United States v. Scarborough, 128 F. 3d 1373, 1378, n. 3 (CA10 1997) (describing a dog that erroneously alerted 4 times out of 19 while working for the postal service and 8% of the time over its entire career); United States v. Limares, 269 F. 3d 794, 797 (CA7 2001) (accepting as reliable a dog that gave false positives between 7% and 38% of the time); Laime v. State, 347 Ark. 142, 159, 60 S. W. 3d 464, 476 (2001) (speaking of a dog that made between 10 and 50 errors); United States v. $242, 484.00, 351 F. 3d 499, 511 (CA11 2003) (noting that because as much as 80% of all currency in circulation contains drug residue, a dog alert "is of little value"), vacated on other grounds by rehearing en banc, 357 F. 3d 1225 (CA11 2004); United States v. Carr, 25 F. 3d 1194, 1214-1217 (CA3 1994) (Becker, J., concurring in part and dissenting in part) ("[A] substantial portion of United States currency . . . is tainted with sufficient traces of controlled substances to cause a trained canine to alert to their presence"). Indeed, a study cited by Illinois in this case for the proposition that dog sniffs are "generally reliable" shows that dogs in artificial testing situations return false positives anywhere from 12.5% to 60% of the time, depending on the length of the search. See Reply Brief for Petitioner 13; Federal Aviation Admin., K. Garner et al., Duty Cycle of the Detector Dog: A Baseline Study 12 (Apr. 2001) (prepared by Auburn U. Inst, for Biological Detection Systems). In practical terms, the evidence is clear that the dog that alerts hundreds of times will be wrong dozens of times.

Once the dog's fallibility is recognized, however, that ends the justification claimed in Place for treating the sniff as sui generis under the Fourth Amendment: the sniff alert does not necessarily signal hidden contraband, and opening the container or enclosed space whose emanations the dog has sensed will not necessarily reveal contraband or any other evidence of crime. This is not, of course, to deny that a dog's reaction may provide reasonable suspicion, or probable cause, to search the container or enclosure; the Fourth Amendment does not demand certainty of success to justify a search for evidence or contraband. The point is simply that the sniff and alert cannot claim the certainty that Place assumed, both in treating the deliberate use of sniffing dogs as sui generis and then taking that characterization as a reason to say they are not searches subject to Fourth Amendment scrutiny. And when that aura of uniqueness disappears, there is no basis in Place's reasoning, and no good reason otherwise, to ignore the actual function that dog sniffs perform. They are conducted to obtain information about the contents of private spaces beyond anything that human senses could perceive, even when conventionally enhanced. The information is not provided by independent third parties beyond the reach of constitutional limitations, but gathered by the government's own officers in order to justify searches of the traditional sort, which may or may not reveal evidence of crime but will disclose anything meant to be kept private in the area searched. Thus in practice the government's use of a trained narcotics dog functions as a limited search to reveal undisclosed facts about private enclosures, to be used to justify a further and complete search of the enclosed area. And given the fallibility of the dog, the sniff is the first step in a process that may disclose "intimate details" without revealing contraband, just as a thermal-imaging device might do, as described in Kyllo v. United States, 533 U.S. 27 (2001).

It makes sense, then, to treat a sniff as the search that it amounts to in practice, and to rely on the body of our Fourth Amendment cases, including Kyllo, in deciding whether such a search is reasonable. As a general proposition, using a dog to sniff for drugs is subject to the rule that the object of enforcing criminal laws does not, without more, justify suspicionless Fourth Amendment intrusions. See Indianapolis v. Edmond, 531 U.S. 32, 41-42 (2000). Since the police claim to have had no particular suspicion that Caballes was violating any drug law, this sniff search must stand or fall on its being ancillary to the traffic stop that led up to it. It is true that the police had probable cause to stop the car for an offense committed in the officer's presence, which Caballes concedes could have justified his arrest. See Brief for Respondent 31. There is no occasion to consider authority incident to arrest, however, see Knowles v. Iowa, 525 U.S. 113 (1998), for the police did nothing more than detain Caballes long enough to check his record and write a ticket. As a consequence, the reasonableness of the search must be assessed in relation to the actual delay the police chose to impose, and as JUSTICE GINSBURG points out in her opinion, post, at 419-420, the Fourth Amendment consequences of stopping for a traffic citation are settled law.

In Berkemer v. McCarty, 468 U.S. 420, 439-440 (1984), followed in Knowles, supra, at 117, we held that the analogue of the common traffic stop was the limited detention for investigation authorized by Terry v. Ohio, 392 U.S. 1 (1968). While Terry authorized a restricted incidental search for weapons when reasonable suspicion warrants such a safety measure, id., at 25-26, the Court took care to keep a Terry stop from automatically becoming a foot in the door for all investigatory purposes; the permissible intrusion was bounded by the justification for the detention, id., at 29-30. Although facts disclosed by enquiry within this limit might give grounds to go further, the government could not otherwise take advantage of a suspect's immobility to search for evidence unrelated to the reason for the detention. That has to be the rule unless Terry is going to become an open sesame for general searches, and that rule requires holding that the police do not have reasonable grounds to conduct sniff searches for drugs simply because they have stopped someone to receive a ticket for a highway offense. Since the police had no indication of illegal activity beyond the speed of the car in this case, the sniff search should be held unreasonable under the Fourth Amendment and its fruits should be suppressed.

Nothing in the case relied upon by the Court, United States v. Jacobsen, 466 U.S. 109 (1984), unsettled the limit of reasonable enquiry adopted in Terry. In Jacobsen, the Court found that no Fourth Amendment search occurred when federal agents analyzed powder they had already lawfully obtained. The Court noted that because the test could only reveal whether the powder was cocaine, the owner had no legitimate privacy interest at stake. 466 U.S., at 123. As already explained, however, the use of a sniffing dog in cases like this is significantly different and properly treated as a search that does indeed implicate Fourth Amendment protection.

In Jacobsen, once the powder was analyzed, that was effectively the end of the matter: either the powder was cocaine, a fact the owner had no legitimate interest in concealing, or it was not cocaine, in which case the test revealed nothing about the powder or anything else that was not already legitimately obvious to the police. But in the case of the dog sniff, the dog does not smell the disclosed contraband; it smells a closed container. An affirmative reaction therefore does not identify a substance the police already legitimately possess, but informs the police instead merely of a reasonable chance of finding contraband they have yet to put their hands on. The police will then open the container and discover whatever lies within, be it marijuana or the owner's private papers. Thus, while Jacobsen could rely on the assumption that the enquiry in question would either show with certainty that a known substance was contraband or would reveal nothing more, both the certainty and the limit on disclosure that may follow are missing when the dog sniffs the car.

The Court today does not go so far as to say explicitly that sniff searches by dogs trained to sense contraband always get a free pass under the Fourth Amendment, since it reserves judgment on the constitutional significance of sniffs assumed to be more intrusive than a dog's walk around a stopped car, ante, at 409. For this reason, I do not take the Court's reliance on Jacobsen as actually signaling recognition of a broad authority to conduct suspicionless sniffs for drugs in any parked car, about which JUSTICE GINSBURG is rightly concerned, post, at 422, or on the person of any pedestrian minding his own business on a sidewalk. But the Court's stated reasoning provides no apparent stopping point short of such excesses. For the sake of providing a workable framework to analyze cases on facts like these, which are certain to come along, I would treat the dog sniff as the familiar search it is in fact, subject to scrutiny under the Fourth Amendment.

JUSTICE GINSBURG, with whom JUSTICE SOUTER joins, dissenting.

Illinois State Police Trooper Daniel Gillette stopped Roy Caballes for driving 71 miles per hour in a zone with a posted speed limit of 65 miles per hour. Trooper Craig Graham of the Drug Interdiction Team heard on the radio that Trooper Gillette was making a traffic stop. Although Gillette requested no aid, Graham decided to come to the scene to conduct a dog sniff. Gillette informed Caballes that he was speeding and asked for the usual documents — driver's license, car registration, and proof of insurance. Caballes promptly provided the requested documents but refused to consent to a search of his vehicle. After calling his dispatcher to check on the validity of Caballes' license and for outstanding warrants, Gillette returned to his vehicle to write Caballes a warning ticket. Interrupted by a radio call on an unrelated matter, Gillette was still writing the ticket when Trooper Graham arrived with his drug-detection dog. Graham walked the dog around the car, the dog alerted at Caballes' trunk, and, after opening the trunk, the troopers found marijuana. 207 Ill. 2d 504, 506-507, 802 N.E. 2d 202, 203 (2003).

The Supreme Court of Illinois held that the drug evidence should have been suppressed. Id., at 506, 802 N. E. 2d, at 202. Adhering to its decision in People v. Cox, 202 Ill. 2d 462, 782 N. E. 2d 275 (2002), the court employed a two-part test taken from Terry v. Ohio, 392 U.S. 1 (1968), to determine the overall reasonableness of the stop. 207 Ill. 2d, at 508, 802 N. E. 2d, at 204. The court asked first "whether the officer's action was justified at its inception," and second "whether it was reasonably related in scope to the circumstances which justified the interference in the first place." Ibid. (quoting People v. Brownlee, 186 Ill. 2d 501, 518-519, 713 N. E. 2d 556, 565 (1999) (in turn quoting Terry, 392 U.S., at 19-20)). "[I]t is undisputed," the court observed, "that the traffic stop was properly initiated"; thus, the dispositive inquiry trained on the "second part of the Terry test," in which "[t]he State bears the burden of establishing that the conduct remained within the scope of the stop." 207 Ill. 2d, at 509, 802 N. E. 2d, at 204.

The court concluded that the State failed to offer sufficient justification for the canine sniff: "The police did not detect the odor of marijuana in the car or note any other evidence suggesting the presence of illegal drugs." Ibid. Lacking "specific and articulable facts" supporting the canine sniff, ibid. (quoting Cox, 202 Ill. 2d, at 470-471, 782 N. E. 2d, at 281), the court ruled, "the police impermissibly broadened the scope of the traffic stop in this case into a drug investigation." 207 Ill. 2d, at 509, 802 N. E. 2d, at 204. I would affirm the Illinois Supreme Court's judgment and hold that the drug sniff violated the Fourth Amendment.

In Terry v. Ohio, the Court upheld the stop and subsequent frisk of an individual based on an officer's observation of suspicious behavior and his reasonable belief that the suspect was armed. See 392 U.S., at 27-28. In a Terry-type investigatory stop, "the officer's action [must be] justified at its inception, and . . . reasonably related in scope to the circumstances which justified the interference in the first place." Id., at 20. In applying Terry, the Court has several times indicated that the limitation on "scope" is not confined to the duration of the seizure; it also encompasses the manner in which the seizure is conducted. See, e.g., Hiibel v. Sixth Judicial Dist. Court of Nev., Humboldt Cty., 542 U.S. 177, 188 (2004) (an officer's request that an individual identify himself "has an immediate relation to the purpose, rationale, and practical demands of a Terry stop"); United States v. Hensley, 469 U.S. 221, 235 (1985) (examining, under Terry, both "the length and intrusiveness of the stop and detention"); Florida v. Royer, 460 U.S. 491, 500 (1983) (plurality opinion) ("[A]n investigative detention must be temporary and last no longer than is necessary to effectuate the purpose of the stop [and] the investigative methods employed should be the least intrusive means reasonably available to verify or dispel the officer's suspicion. . . .").

"A routine traffic stop," the Court has observed, "is a relatively brief encounter and `is more analogous to a socalled Terry stop . . . than to a formal arrest."' Knowles v. Iowa, 525 U.S. 113, 117 (1998) (quoting Berkemer v. McCarty, 468 U.S. 420, 439 (1984)); see also ante, at 415 (SOUTER, J., dissenting) (The government may not "take advantage of a suspect's immobility to search for evidence unrelated to the reason for the detention."). I would apply Terry's reasonable-relation test, as the Illinois Supreme Court did, to determine whether the canine sniff impermissibly expanded the scope of the initially valid seizure of Caballes. It is hardly dispositive that the dog sniff in this case may not have lengthened the duration of the stop. Cf. ante, at 407 ("A seizure . . . can become unlawful if it is prolonged beyond the time reasonably required to complete [the initial] mission."). Terry, it merits repetition, instructs that any investigation must be "reasonably related in scope to the circumstances which justified the interference in the first place." 392 U.S., at 20 (emphasis added). The unwarranted and nonconsensual expansion of the seizure here from a routine traffic stop to a drug investigation broadened the scope of the investigation in a manner that, in my judgment, runs afoul of the Fourth Amendment.

The Court rejects the Illinois Supreme Court's judgment and, implicitly, the application of Terry to a traffic stop converted, by calling in a dog, to a drug search. The Court so rules, holding that a dog sniff does not render a seizure that is reasonable in time unreasonable in scope. Ante, at 408. Dog sniffs that detect only the possession of contraband may be employed without offense to the Fourth Amendment, the Court reasons, because they reveal no lawful activity and hence disturb no legitimate expectation of privacy. Ante, at 408-409.

In my view, the Court diminishes the Fourth Amendment's force by abandoning the second Terry inquiry (was the police action "reasonably related in scope to the circumstances [justifiying] the [initial] interference"). 392 U.S., at 20. A drug-detection dog is an intimidating animal. Cf. United States v. Williams, 356 F. 3d 1268, 1276 (CA10 2004) (McKay, J., dissenting) ("drug dogs are not lap dogs"). Injecting such an animal into a routine traffic stop changes the character of the encounter between the police and the motorist. The stop becomes broader, more adversarial, and (in at least some cases) longer. Caballes — who, as far as Troopers Gillette and Graham knew, was guilty solely of driving six miles per hour over the speed limit — was exposed to the embarrassment and intimidation of being investigated, on a public thoroughfare, for drugs. Even if the drug sniff is not characterized as a Fourth Amendment "search," cf. Indianapolis v. Edmond, 531 U.S. 32, 40 (2000); United States v. Place, 462 U.S. 696, 707 (1983), the sniff surely broadened the scope of the traffic-violation-related seizure.

The Court has never removed police action from Fourth Amendment control on the ground that the action is well calculated to apprehend the guilty. See, e.g., United States v. Karo, 468 U.S. 705, 717 (1984) ( Fourth Amendment warrant requirement applies to police monitoring of a beeper in a house even if "the facts [justify] believing that a crime is being or will be committed and that monitoring the beeper wherever it goes is likely to produce evidence of criminal activity."); see also Minnesota v. Carter, 525 U.S. 83, 110 (1998) (GINSBURG, J., dissenting) ("Fourth Amendment protection, reserved for the innocent only, would have little force in regulating police behavior toward either the innocent or the guilty."). Under today's decision, every traffic stop could become an occasion to call in the dogs, to the distress and embarrassment of the law-abiding population.

The Illinois Supreme Court, it seems to me, correctly apprehended the danger in allowing the police to search for contraband despite the absence of cause to suspect its presence. Today's decision, in contrast, clears the way for suspicionless, dog-accompanied drug sweeps of parked cars along sidewalks and in parking lots. Compare, e.g., United States v. Ludwig, 10 F. 3d 1523, 1526-1527 (CA10 1993) (upholding a search based on a canine drug sniff of a parked car in a motel parking lot conducted without particular suspicion), with United States v. Quinn, 815 F. 2d 153, 159 (CA1 1987) (officers must have reasonable suspicion that a car contains narcotics at the moment a dog sniff is performed), and Place, 462 U.S., at 706-707 ( Fourth Amendment not violated by a dog sniff of a piece of luggage that was seized, pre-sniff, based on suspicion of drugs). Nor would motorists have constitutional grounds for complaint should police with dogs, stationed at long traffic lights, circle cars waiting for the red signal to turn green.

Today's decision also undermines this Court's situation-sensitive balancing of Fourth Amendment interests in other contexts. For example, in Bond v. United States, 529 U.S. 334, 338-339 (2000), the Court held that a bus passenger had an expectation of privacy in a bag placed in an overhead bin and that a police officer's physical manipulation of the bag constituted an illegal search. If canine drug sniffs are entirely exempt from Fourth Amendment inspection, a sniff could substitute for an officer's request to a bus passenger for permission to search his bag, with this significant difference: The passenger would not have the option to say "No."

The dog sniff in this case, it bears emphasis, was for drug detection only. A dog sniff for explosives, involving security interests not presented here, would be an entirely different matter. Detector dogs are ordinarily trained not as all-purpose sniffers, but for discrete purposes. For example, they may be trained for narcotics detection or for explosives detection or for agricultural products detection. See, e.g., U.S. Customs Border Protection, Canine Enforcement Training Center Training Program Course Descriptions, http://www.cbp.gov/xp/cgov/border_security/canines/training_program.xml (all Internet materials as visited Dec. 16, 2004, and available in Clerk of Court's case file) (describing Customs training courses in narcotics detection); Transportation Security Administration, Canine and Explosives Program, http://www.tsa.gov/public/display?theme=32 (describing Transportation Security Administration's explosives detection canine program); U.S. Dept. of Agriculture, Animal and Plant Health Inspection Service, USDA's Detector Dogs: Protecting American Agriculture (Oct. 2001), available at http://www.aphis.usda.gov/oa/pubs/detdogs.pdf (describing USDA Beagle Brigade detector dogs trained to detect prohibited fruits, plants, and meat); see also Jennings, Origins and History of Security and Detector Dogs, in Canine Sports Medicine and Surgery 16, 18-19 (M. Bloomberg, J. Dee, R. Taylor eds. 1998) (describing narcotics-detector dogs used by Border Patrol and Customs, and bomb detector dogs used by the Federal Aviation Administration and the Secret Service, but noting the possibility in some circumstances of cross training dogs for multiple tasks); S. Chapman, Police Dogs in North America 64, 70-79 (1990) (describing narcotics and explosives-detection dogs and noting the possibility of cross training). There is no indication in this case that the dog accompanying Trooper Graham was trained for anything other than drug detection. See 207 Ill. 2d, at 507, 802 N. E. 2d, at 203 ("Trooper Graham arrived with his drug-detection dog. . . ."); Brief for Petitioner 3 ("Trooper Graham arrived with a drug-detection dog. . . .").

This Court has distinguished between the general interest in crime control and more immediate threats to public safety. In Michigan Dept. of State Police v. Sitz, 496 U. S. 444 (1990), this Court upheld the use of a sobriety traffic check-point. Balancing the State's interest in preventing drunk driving, the extent to which that could be accomplished through the checkpoint program, and the degree of intrusion the stops involved, the Court determined that the State's checkpoint program was consistent with the Fourth Amendment. Id., at 455. Ten years after Sitz, in Indianapolis v. Edmond, 531 U.S. 32, this Court held that a drug interdiction checkpoint violated the Fourth Amendment. Despite the illegal narcotics traffic that the Nation is struggling to stem, the Court explained, a "general interest in crime control" did not justify the stops. Id., at 43-44 (internal quotation marks omitted). The Court distinguished the sobriety checkpoints in Sitz on the ground that those checkpoints were designed to eliminate an "immediate, vehicle-bound threat to life and limb." 531 U.S., at 43.

The use of bomb-detection dogs to check vehicles for explosives without doubt has a closer kinship to the sobriety checkpoints in Sitz than to the drug checkpoints in Edmond. As the Court observed in Edmond: "[T]he Fourth Amendment would almost certainly permit an appropriately tailored roadblock set up to thwart an imminent terrorist attack. . . ." 531 U.S., at 44. Even if the Court were to change course and characterize a dog sniff as an independent Fourth Amendment search, see ante, p. 410 (SOUTER, J., dissenting), the immediate, present danger of explosives would likely justify a bomb sniff under the special needs doctrine. See, e.g., ante, at 417, n. 7 (SOUTER, J., dissenting); Griffin v. Wisconsin, 483 U.S. 868, 873 (1987) (permitting exceptions to the warrant and probable-cause requirements for a search when "special needs, beyond the normal need for law enforcement," make those requirements impracticable (quoting New Jersey v. T. L. O., 469 U.S. 325, 351 (1985) (Blackmun, J., concurring in judgment))).

* * *

For the reasons stated, I would hold that the police violated Caballes' Fourth Amendment rights when, without cause to suspect wrongdoing, they conducted a dog sniff of his vehicle. I would therefore affirm the judgment of the Illinois Supreme Court.