Florida v. Harris

237

Syllabus

certiorari to the supreme court of Ćorida No. 11–817. Argued October 31, 2012—Decided February 19, 2013

Officer Wheetley pulled over respondent Harris for a routine traffic stop.

Observing Harris’s nervousness and an open beer can, Wheetley sought consent to search Harris’s truck. When Harris refused, Wheetley exe­ cuted a sniff test with his trained narcotics dog, Aldo. The dog alerted at the driver’s-side door handle, leading Wheetley to conclude that he had probable cause for a search. That search turned up nothing Aldo was trained to detect, but did reveal pseudoephedrine and other ingredi­ ents for manufacturing methamphetamine. Harris was arrested and charged with illegal possession of those ingredients. In a subsequent stop while Harris was out on bail, Aldo again alerted on Harris’s truck but nothing of interest was found. At a suppression hearing, Wheetley testified about his and Aldo’s extensive training in drug detection.

Harris’s attorney did not contest the quality of that training, focusing instead on Aldo’s certification and performance in the field, particularly in the two stops of Harris’s truck. The trial court denied the motion to suppress, but the Florida Supreme Court reversed. It held that a wide array of evidence was always necessary to establish probable cause, in­ cluding field performance records showing how many times the dog has falsely alerted. If an officer like Wheetley failed to keep such records, he could never have probable cause to think the dog a reliable indicator of drugs.

Held: Because training and testing records supported Aldo’s reliability in detecting drugs and Harris failed to undermine that evidence, Wheetley had probable cause to search Harris’s truck. Pp. 243–250.

(a) In testing whether an officer has probable cause to conduct a search, all that is required is the kind of “fair probability” on which “reasonable and prudent [people] act.” Illinois v. Gates, 462 U. S. 213,

235. To evaluate whether the State has met this practical and common­

sensical standard, this Court has consistently looked to the totality of the circumstances and rejected rigid rules, bright-line tests, and mecha­ nistic inquiries. Ibid.

The Florida Supreme Court flouted this established approach by cre­ ating a strict evidentiary checklist to assess a drug-detection dog’s relia­ bility. Requiring the State to introduce comprehensive documentation of the dog’s prior hits and misses in the field, and holding that absent

238

Syllabus

field records will preclude a finding of probable cause no matter how much other proof the State offers, is the antithesis of a totality-of-the­ circumstances approach. This is made worse by the State Supreme Court’s treatment of field performance records as the evidentiary gold standard when, in fact, such data may not capture a dog’s false negatives or may markedly overstate a dog’s false positives. Such inaccuracies do not taint records of a dog’s performance in standard training and certification settings, making that performance a better measure of a dog’s reliability. Field records may sometimes be relevant, but the court should evaluate all the evidence, and should not prescribe an in­ flexible set of requirements.

Under the correct approach, a probable-cause hearing focusing on a dog’s alert should proceed much like any other, with the court allowing the parties to make their best case and evaluating the totality of the circumstances. If the State has produced proof from controlled settings that a dog performs reliably in detecting drugs, and the defendant has not contested that showing, the court should find probable cause. But a defendant must have an opportunity to challenge such evidence of a dog’s reliability, whether by cross-examining the testifying officer or by introducing his own fact or expert witnesses. The defendant may con­ test training or testing standards as flawed or too lax, or raise an issue regarding the particular alert. The court should then consider all the evidence and apply the usual test for probable cause—whether all the facts surrounding the alert, viewed through the lens of common sense, would make a reasonably prudent person think that a search would re­ veal contraband or evidence of a crime. Pp. 243–248.

(b) The record in this case amply supported the trial court’s determi­ nation that Aldo’s alert gave Wheetley probable cause to search the truck. The State introduced substantial evidence of Aldo’s training and his proficiency in finding drugs. Harris declined to challenge any as­ pect of that training or testing in the trial court, and the Court does not consider such arguments when they are presented for the first time in this Court. Harris principally relied below on Wheetley’s failure to find any substance that Aldo was trained to detect. That infers too much from the failure of a particular alert to lead to drugs, and did not rebut the State’s evidence from recent training and testing. Pp. 248–250.

71 So. 3d 756, reversed.

Kagan, J., delivered the opinion for a unanimous Court.

Gregory G. Garre argued the cause for petitioner. With him on the briefs were Pamela Jo Bondi, Attorney Gen­ eral of Florida, Carolyn M. Snurkowski, Associate Deputy

Cite as: 568 U. S. 237 (2013)

239

Counsel

Attorney General, Robert J. Krauss, Chief-Assistant Attor­ ney General, and Susan M. Shanahan, Assistant Attorney General.

Joseph R. Palmore argued the cause for the United States as amicus curiae in support of petitioner. With him on the brief were Solicitor General Verrilli, Assistant Attorney General Breuer, and Deputy Solicitor General Dreeben.

Glen P. Gifford argued the cause and filed a brief for respondent.*

*Briefs of amici curiae urging reversal were filed for the Common­ wealth of Virginia et al. by Kenneth T. Cuccinelli II, Attorney General of Virginia, E. Duncan Getchell, Jr., Solicitor General, Charles E. James, Jr., Chief Deputy Attorney General, Wesley G. Russell, Jr., Deputy Attorney General, and Michael H. Brady, Assistant Attorney General, and by offi­ cials for their respective jurisdictions as follows: Luther Strange, Attorney General of Alabama, Tom Horne, Attorney General of Arizona, John W.

Suthers, Attorney General of Colorado, Joseph R. Biden III, Attorney General of Delaware, Leonardo M. Rapadas, Attorney General of Guam, Lawrence G. Wasden, Attorney General of Idaho, Lisa Madigan, Attorney General of Illinois, Gregory F. Zoeller, Attorney General of Indiana, Derek Schmidt, Attorney General of Kansas, and John Campbell, Chief Deputy Attorney General, William J. Schneider, Attorney General of Maine, Bill Schuette, Attorney General of Michigan, Chris Koster, Attorney General of Missouri, Jon Bruning, Attorney General of Nebraska, Jeffrey S. Chiesa, Attorney General of New Jersey, Gary K. King, Attorney General of New Mexico, Wayne Stenehjem, Attorney General of North Dakota, E. Scott Pruitt, Attorney General of Oklahoma, John R. Kroger, Attorney General of Oregon, Linda L. Kelly, Attorney General of Pennsylvania, Guillermo Somoza-Colombani, Attorney General of Puerto Rico, and Luis R. Román Negrón, Solicitor General, Greg Abbott, Attorney General of Texas, Mark

General of Vermont, Robert M. McKenna, Attorney General of Washing­ ton, J. B. Van Hollen, Attorney General of Wisconsin, and Gregory A.

Phillips, Attorney General of Wyoming; and for the National Police Ca­ nine Association et al. by Arthur T. Daus III.

Briefs of amici curiae urging affirmance were filed for the Electronic Privacy Information Center by Marc Rotenberg; for the Institute for Jus­ tice by William H. Mellor, Scott G. Bullock, Darpana M. Sheth, and Rob­ ert P. Frommer; for the National Association of Criminal Defense Law­ yers et al. by Danielle Spinelli, Annie L. Owens, Jonathan D. Hacker,

240

Opinion of the Court Justice Kagan delivered the opinion of the Court.

In this case, we consider how a court should determine if the “alert” of a drug-detection dog during a traffic stop pro­ vides probable cause to search a vehicle. The Florida Su­ preme Court held that the State must in every case present an exhaustive set of records, including a log of the dog’s per­ formance in the field, to establish the dog’s reliability. See 71 So. 3d 756, 775 (2011). We think that demand inconsist­ ent with the “flexible, common-sense standard” of probable cause. Illinois v. Gates, 462 U. S. 213, 239 (1983).

William Wheetley is a K–9 Officer in the Liberty County, Florida Sheriff ’s Office. On June 24, 2006, he was on a rou­ tine patrol with Aldo, a German shepherd trained to detect certain narcotics (methamphetamine, marijuana, cocaine, heroin, and ecstasy). Wheetley pulled over respondent Clayton Harris’s truck because it had an expired license plate. On approaching the driver’s-side door, Wheetley saw that Harris was “visibly nervous,” unable to sit still, shaking, and breathing rapidly. App. 62. Wheetley also noticed an open can of beer in the truck’s cup holder. Ibid. Wheetley asked Harris for consent to search the truck, but Harris re­ fused. At that point, Wheetley retrieved Aldo from the pa­ trol car and walked him around Harris’s truck for a “free air sniff.” Id., at 63. Aldo alerted at the driver’s-side door handle—signaling, through a distinctive set of behaviors, that he smelled drugs there.

Wheetley concluded, based principally on Aldo’s alert, that he had probable cause to search the truck. His search did not turn up any of the drugs Aldo was trained

Mason C. Clutter, Steven R. Shapiro, Ezekiel R. Edwards, Michael Uffer­ man, Sonya Rudenstine, Randall C. Marshall, and Maria Kayanan; for The Rutherford Institute by Anand Agneshwar, Lisa S. Blatt, Carl Nadler, John W. Whitehead, Rita M. Dunaway, and Charles I. Lugosi; and for Fourth Amendment Scholars by Leslie A. Shoebotham.

Cite as: 568 U. S. 237 (2013)

241

Opinion of the Court

to detect. But it did reveal 200 loose pseudoephedrine pills, 8,000 matches, a bottle of hydrochloric acid, two containers of antifreeze, and a coffee filter full of iodine crystals—all ingredients for making methamphetamine.

Wheetley accordingly arrested Harris, who admitted after proper Miranda warnings that he routinely “cooked” meth­ amphetamine at his house and could not go “more than a few days without using” it. Id., at 68. The State charged Har­ ris with possessing pseudoephedrine for use in manufactur­ ing methamphetamine.

While out on bail, Harris had another run-in with Wheet­ ley and Aldo. This time, Wheetley pulled Harris over for a broken brake light. Aldo again sniffed the truck’s exterior, and again alerted at the driver’s-side door handle. Wheet­ ley once more searched the truck, but on this occasion discov­ ered nothing of interest.

Harris moved to suppress the evidence found in his truck on the ground that Aldo’s alert had not given Wheetley prob­ able cause for a search. At the hearing on that motion, Wheetley testified about both his and Aldo’s training in drug detection. See id., at 52–82. In 2004, Wheetley (and a dif­ ferent dog) completed a 160-hour course in narcotics detec­ tion offered by the Dothan, Alabama Police Department, while Aldo (and a different handler) completed a similar, 120­ hour course given by the Apopka, Florida Police Depart­ ment. That same year, Aldo received a one-year certifi­ cation from Drug Beat, a private company that specializes in testing and certifying K–9 dogs. Wheetley and Aldo teamed up in 2005 and went through another, 40-hour re­ fresher course in Dothan together. They also did four hours of training exercises each week to maintain their skills.

Wheetley would hide drugs in certain vehicles or buildings while leaving others “blank” to determine whether Aldo alerted at the right places. Id., at 57. According to Wheet­ ley, Aldo’s performance in those exercises was “really good.” Id., at 60. The State introduced “Monthly Canine Detection

242

Opinion of the Court

Training Logs” consistent with that testimony: They showed that Aldo always found hidden drugs and that he performed “satisfactorily” (the higher of two possible assessments) on each day of training. Id., at 109–116.

On cross-examination, Harris’s attorney chose not to con­ test the quality of Aldo’s or Wheetley’s training. She fo­ cused instead on Aldo’s certification and his performance in the field, particularly the two stops of Harris’s truck.

Wheetley conceded that the certification (which, he noted, Florida law did not require) had expired the year before he pulled Harris over. See id., at 70–71. Wheetley also ac­ knowledged that he did not keep complete records of Aldo’s performance in traffic stops or other field work; instead, he maintained records only of alerts resulting in arrests. See id., at 71–72, 74. But Wheetley defended Aldo’s two alerts to Harris’s seemingly narcotics-free truck: According to Wheetley, Harris probably transferred the odor of metham­ phetamine to the door handle, and Aldo responded to that “residual odor.” Id., at 80.

The trial court concluded that Wheetley had probable cause to search Harris’s truck and so denied the motion to suppress. Harris then entered a no-contest plea while re­ serving the right to appeal the trial court’s ruling. An in­ termediate state court summarily affirmed. 989 So. 2d 1214, 1215 (Fla. App. 2008) (per curiam).

The Florida Supreme Court reversed, holding that Wheet­ ley lacked probable cause to search Harris’s vehicle under the Fourth Amendment. “[W]hen a dog alerts,” the court wrote, “the fact that the dog has been trained and certified is simply not enough to establish probable cause.” 71 So.

3d, at 767. To demonstrate a dog’s reliability, the State

“[T]he State must present . . . the dog’s training and certification records, an explanation of the meaning of the particular training and certification, field perform­ nance records (including any unverified alerts), and evi­

Cite as: 568 U. S. 237 (2013)

243

Opinion of the Court

dence concerning the experience and training of the of­ ficer handling the dog, as well as any other objective evidence known to the officer about the dog’s reliability.” Id., at 775.

The court particularly stressed the need for “evidence of the dog’s performance history,” including records showing “how often the dog has alerted in the field without illegal contra­ band having been found.” Id., at 769. That data, the court stated, could help to expose such problems as a handler’s tendency (conscious or not) to “cue [a] dog to alert” and “a dog’s inability to distinguish between residual odors and ac­ tual drugs.” Id., at 769, 774. Accordingly, an officer like Wheetley who did not keep full records of his dog’s field per­ formance could never have the requisite cause to think “that the dog is a reliable indicator of drugs.” Id., at 773.

Chief Justice Canady dissented, maintaining that the ma­ jority’s “elaborate and inflexible evidentiary requirements” went beyond the demands of probable cause. Id., at 775.

He would have affirmed the trial court’s ruling on the strength of Aldo’s training history and Harris’s “fail[ure] to present any evidence challenging” it. Id., at 776.

We granted certiorari, 566 U. S. 904 (2012), and now reverse.

A police officer has probable cause to conduct a search when “the facts available to [him] would ‘warrant a [person] of reasonable caution in the belief ’ ” that contraband or evi­ dence of a crime is present. Texas v. Brown, 460 U. S. 730, 742 (1983) (plurality opinion) (quoting Carroll v. United States, 267 U. S. 132, 162 (1925)); see Safford Unified School Dist. #1 v. Redding, 557 U. S. 364, 370–371 (2009). The test for probable cause is not reducible to “precise definition or quantification.” Maryland v. Pringle, 540 U. S. 366, 371 (2003). “Finely tuned standards such as proof beyond a reasonable doubt or by a preponderance of the evidence . . .

244

Opinion of the Court have no place in the [probable-cause] decision.” Gates, 462

U. S., at 235. All we have required is the kind of “fair proba­

bility” on which “reasonable and prudent [people,] not legal technicians, act.” Id., at 238, 231 (internal quotation marks omitted).

In evaluating whether the State has met this practical and common-sensical standard, we have consistently looked to the totality of the circumstances. See, e. g., Pringle, 540

U. S., at 371; Gates, 462 U. S., at 232; Brinegar v. United

States, 338 U. S. 160, 176 (1949). We have rejected rigid rules, bright-line tests, and mechanistic inquiries in favor of a more flexible, all-things-considered approach. In Gates, for example, we abandoned our old test for assessing the re­ liability of informants’ tips because it had devolved into a “complex superstructure of evidentiary and analytical rules,” any one of which, if not complied with, would derail a finding of probable cause. 462 U. S., at 235. We lamented the de­ velopment of a list of “inflexible, independent requirements applicable in every case.” Id., at 230, n. 6. Probable cause, we emphasized, is “a fluid concept—turning on the assess­ ment of probabilities in particular factual contexts—not readily, or even usefully, reduced to a neat set of legal rules.” Id., at 232.

The Florida Supreme Court flouted this established approach to determining probable cause. To assess the reliability of a drug-detection dog, the court created a strict evidentiary checklist, whose every item the State must tick off.1 Most prominently, an alert cannot establish

1 By the time of oral argument in this case, even Harris declined to defend the idea that the Fourth Amendment compels the State to produce each item of evidence the Florida Supreme Court enumerated. See Tr. of Oral Arg. 29–30 (“I don’t believe the Constitution requires [that list]”).

Harris instead argued that the court’s decision, although “look[ing] rather didactic,” in fact did not impose any such requirement. Id., at 29; see id., at 31 (“[I]t’s not a specific recipe that can’t be deviated from”). But in reading the decision below as establishing a mandatory checklist, we

Cite as: 568 U. S. 237 (2013)

245

Opinion of the Court

probable cause under the Florida court’s decision unless the State introduces comprehensive documentation of the dog’s prior “hits” and “misses” in the field.

(One wonders how the court would apply its test to a rookie dog.) No matter how much other proof the State offers of the dog’s reliability, the absent field performance records will preclude a finding of probable cause. That is the antithesis of a totality-of-the­ circumstances analysis. It is, indeed, the very thing we criticized in Gates when we overhauled our method for assessing the trustworthiness of an informant’s tip. A gap as to any one matter, we explained, should not sink the State’s case; rather, that “deficiency . . . may be compensated for, in determining the overall reliability of a tip, by a strong showing as to . . . other indicia of reliability.” Id., at

233. So too here, a finding of a drug-detection dog’s reliabil­

ity cannot depend on the State’s satisfaction of multiple, independent evidentiary requirements. No more for dogs than for human informants is such an inflexible checklist the way to prove reliability, and thus establish probable cause.

Making matters worse, the decision below treats records of a dog’s field performance as the gold standard in evidence, when in most cases they have relatively limited import. Er­ rors may abound in such records. If a dog on patrol fails to alert to a car containing drugs, the mistake usually will go undetected because the officer will not initiate a search.

Field data thus may not capture a dog’s false negatives.

Conversely (and more relevant here), if the dog alerts to a car in which the officer finds no narcotics, the dog may not have made a mistake at all. The dog may have detected substances that were too well hidden or present in quantities too small for the officer to locate. Or the dog may have smelled the residual odor of drugs previously in the vehicle

do no more than take the court at its (oft-repeated) word. See, e. g., 71 So. 3d 756, 758, 759, 771, 775 (Fla. 2011) (holding that the State “must” present the itemized evidence).

246

Opinion of the Court

or on the driver’s person.2 Field data thus may markedly overstate a dog’s real false positives. By contrast, those in­ accuracies—in either direction—do not taint records of a dog’s performance in standard training and certification set­ tings. There, the designers of an assessment know where drugs are hidden and where they are not—and so where a dog should alert and where he should not. The better meas­ ure of a dog’s reliability thus comes away from the field, in controlled testing environments.3

For that reason, evidence of a dog’s satisfactory perform­ ance in a certification or training program can itself provide sufficient reason to trust his alert. If a bona fide organiza­

2 See U. S. Dept. of Army, Military Working Dog Program 30 (Pamphlet 190–12, 1993) (“The odor of a substance may be present in enough concen­ tration to cause the dog to respond even after the substance has been removed. Therefore, when a detector dog responds and no drug or explo­ sive is found, do not assume the dog has made an error”); S. Bryson, Police Dog Tactics 257 (2d ed. 2000) (“Four skiers toke up in the parking lot before going up the mountain. Five minutes later a narcotic detector dog alerts to the car. There is no dope inside. However, the dog has per­ formed correctly”). The Florida Supreme Court treated a dog’s response to residual odor as an error, referring to the “inability to distinguish be­ tween [such] odors and actual drugs” as a “facto[r] that call[s] into question Aldo’s reliability.” 71 So. 3d, at 773–774; see supra, at 243. But that statement reflects a misunderstanding. A detection dog recognizes an odor, not a drug, and should alert whenever the scent is present, even if the substance is gone ( just as a police officer’s much inferior nose detects the odor of marijuana for some time after a joint has been smoked). In the usual case, the mere chance that the substance might no longer be at the location does not matter; a well-trained dog’s alert establishes a fair probability—all that is required for probable cause—that either drugs or evidence of a drug crime (like the precursor chemicals in Harris’s truck) will be found.

3 See K. Furton, J. Greb, & H. Holness, Florida Int’l Univ., The Scientific Working Group on Dog and Orthogonal Detector Guidelines 1, 61–62, 66 (2010) (recommending as a “best practice” that a dog’s reliability should be assessed based on “the results of certification and proficiency assess­ ments,” because in those “procedure[s] you should know whether you have a false positive,” unlike in “most operational situations”).

Cite as: 568 U. S. 237 (2013)

247

Opinion of the Court

tion has certified a dog after testing his reliability in a con­ trolled setting, a court can presume (subject to any conflict­ ing evidence offered) that the dog’s alert provides probable cause to search. The same is true, even in the absence of formal certification, if the dog has recently and successfully completed a training program that evaluated his proficiency in locating drugs. After all, law enforcement units have their own strong incentive to use effective training and certi­ fication programs, because only accurate drug-detection dogs enable officers to locate contraband without incurring unnec­ essary risks or wasting limited time and resources.

A defendant, however, must have an opportunity to chal­ lenge such evidence of a dog’s reliability, whether by cross- examining the testifying officer or by introducing his own fact or expert witnesses. The defendant, for example, may contest the adequacy of a certification or training program, perhaps asserting that its standards are too lax or its meth­ ods faulty. So too, the defendant may examine how the dog (or handler) performed in the assessments made in those set­ tings. Indeed, evidence of the dog’s (or handler’s) history in the field, although susceptible to the kind of misinterpreta­ tion we have discussed, may sometimes be relevant, as the Solicitor General acknowledged at oral argument. See Tr. of Oral Arg. 23–24 (“[T]he defendant can ask the handler, if the handler is on the stand, about field performance, and then the court can give that answer whatever weight is appro­ priate”). And even assuming a dog is generally reliable, cir­ cumstances surrounding a particular alert may undermine the case for probable cause—if, say, the officer cued the dog (consciously or not), or if the team was working under unfa­ miliar conditions.

In short, a probable-cause hearing focusing on a dog’s alert should proceed much like any other. The court should allow the parties to make their best case, consistent with the usual rules of criminal procedure. And the court should then evaluate the proffered evidence to decide what all the cir­

248

Opinion of the Court

cumstances demonstrate. If the State has produced proof from controlled settings that a dog performs reliably in de­ tecting drugs, and the defendant has not contested that showing, then the court should find probable cause. If, in contrast, the defendant has challenged the State’s case (by disputing the reliability of the dog overall or of a particular alert), then the court should weigh the competing evidence.

In all events, the court should not prescribe, as the Florida Supreme Court did, an inflexible set of evidentiary require­ ments. The question—similar to every inquiry into proba­ ble cause—is whether all the facts surrounding a dog’s alert, viewed through the lens of common sense, would make a rea­ sonably prudent person think that a search would reveal con­ traband or evidence of a crime. A sniff is up to snuff when it meets that test.

And here, Aldo’s did. The record in this case amply sup­ ported the trial court’s determination that Aldo’s alert gave Wheetley probable cause to search Harris’s truck.

The State, as earlier described, introduced substantial evi­ dence of Aldo’s training and his proficiency in finding drugs.

See supra, at 241–242. The State showed that two years before alerting to Harris’s truck, Aldo had successfully com­ pleted a 120-hour program in narcotics detection, and sepa­ rately obtained a certification from an independent company.

And although the certification expired after a year, the Sher­ iff ’s Office required continuing training for Aldo and Wheet­ ley. The two satisfied the requirements of another, 40-hour training program one year prior to the search at issue. And Wheetley worked with Aldo for four hours each week on exercises designed to keep their skills sharp. Wheetley tes­ tified, and written records confirmed, that in those settings Aldo always performed at the highest level.

Harris, as also noted above, declined to challenge in the trial court any aspect of Aldo’s training. See supra, at 242.

To be sure, Harris’s briefs in this Court raise questions about Cite as: 568 U. S. 237 (2013)

249

Opinion of the Court

that training’s adequacy—for example, whether the pro­ grams simulated sufficiently diverse environments and whether they used enough blind testing (in which the han­ dler does not know the location of drugs and so cannot cue the dog). See Brief for Respondent 57–58. Similarly, Har­ ris here queries just how well Aldo performed in controlled testing. See id., at 58. But Harris never voiced those doubts in the trial court, and cannot do so for the first time here. See, e. g., Rugendorf v. United States, 376 U. S. 528, 534 (1964). As the case came to the trial court, Aldo had successfully completed two recent drug-detection courses and maintained his proficiency through weekly training exer­ cises. Viewed alone, that training record—with or without the prior certification—sufficed to establish Aldo’s reliability.

See supra, at 246–248.

And Harris’s cross-examination of Wheetley, which fo­ cused on Aldo’s field performance, failed to rebut the State’s case. Harris principally contended in the trial court that because Wheetley did not find any of the substances Aldo was trained to detect, Aldo’s two alerts must have been false.

See Brief for Respondent 1; App. 77–80. But we have al­ ready described the hazards of inferring too much from the failure of a dog’s alert to lead to drugs, see supra, at 245–246; and here we doubt that Harris’s logic does justice to Aldo’s skills. Harris cooked and used methamphetamine on a regu­ lar basis; so as Wheetley later surmised, Aldo likely responded to odors that Harris had transferred to the driver’s-side door handle of his truck. See supra, at 242. A well-trained drug-detection dog should alert to such odors; his response to them might appear a mistake, but in fact is not. See n. 2, supra. And still more fundamentally, we do not evaluate probable cause in hindsight, based on what a search does or does not turn up. See United States v. Di Re, 332 U. S. 581, 595 (1948). For the reasons already stated, Wheetley had good cause to view Aldo as a reliable detector of drugs.

And no special circumstance here gave Wheetley reason to

250

Opinion of the Court

discount Aldo’s usual dependability or distrust his response to Harris’s truck.

Because training records established Aldo’s reliability in detecting drugs and Harris failed to undermine that show­ ing, we agree with the trial court that Wheetley had proba­ ble cause to search Harris’s truck. We accordingly reverse the judgment of the Florida Supreme Court.

It is so ordered.