OPINION
WILKINSON, Circuit Judge:
We are asked to review David Kelly’s convictions for conspiracy to distribute and possess with intent to distribute drugs and three counts of drug possession with intent to distribute. Kelly’s principal claim is that the police violated the Fourth Amendment by conducting a warrantless search of an automobile parked on the street in front of his residence. He contends that the automobile exception to the warrant requirement does not apply here because the police were armed, had the only two individuals at his residence under arrest, and possessed the key to the ignition— thus eliminating any immediate threat that the vehicle would be driven away.
We reject Kelly’s attempt to create an exception to the automobile exception. The exception, as carefully crafted by the Supreme Court, does not have an exigency requirement apart from the inherent mobility of the automobile. Consequently, if the police have probable cause, the justification to conduct a warrantless search does not vanish once the police have established some degree of control over the automobile. After careful consideration, we also reject Kelly’s other claims and affirm his convictions.
I.
A.
Federal agents and local narcotics investigators conducted a year-long investigation into the drug dealing activities of David Kelly and other co-conspirators. During the investigation, they learned that Kelly had distributed multiple kilograms of cocaine in and around Hampton Roads, Virginia and that his main cocaine supplier was a Hispanic male from New York City. In addition, they learned that Kelly drove several vehicles, including a dark green Lexus sedan.
On September 19, 2006, a federal magistrate issued arrest warrants for Kelly and two of his co-conspirators. The magistrate also issued search warrants for Kelly’s residence and a commercial building he owned. The warrant for his residence did not refer to any of his vehicles.
On September 20, 2006, a police officer conducted surveillance at Kelly’s residence from 5:00 a.m. until 8:00 p.m. in preparation for the execution of the warrants. She spotted three vehicles owned by Kelly and his girlfriend: two Chevrolet sports utility vehicles parked in the lot of the residence and an Infiniti parked along the street. There was no sign, however, of Kelly or the Lexus at any point that day. On the morning of September 21, the officer resumed her surveillance and immediately noticed that Kelly’s Lexus was parked on the street in front of the residence. She also saw Kelly come out of the residence to jump-start the Infiniti, which his girlfriend then drove away.
Around noon of that day, a search team executed the warrant at Kelly’s residence. Once inside, the team arrested Kelly and a man whom they unexpectedly discovered. The man identified himself as Jose Jiminez and told the police that he was from New York City. In his possession was a bag containing travel items, such as underwear, soap, and toothpaste. Based on these facts, the officers immediately suspected that Jiminez was Kelly’s cocaine supplier and that the two of them had arrived the previous night in the Lexus.
After waiving his
Miranda
rights, Kelly was placed in a police cruiser and questioned by the officers. Kelly initially denied that there were any drugs in either the residence or his three vehicles parked outside. But after learning that a K-9 unit was on its way, Kelly nodded his head “yes” when asked if there was cocaine in the vehicles. He did not, however, specify which of the three vehicles contained the cocaine.
Subsequently, the K-9 unit arrived at the residence. A specially trained officer led a drug detection dog around the Lexus, and the dog alerted positively by turning its head and scratching at the driver’s door. Using Kelly’s car keys to open the vehicle, the officers searched the passenger compartment but did not find any drugs. They then searched the trunk and discovered a backpack containing five kilograms of cocaine and 856 tablets of ecstasy. The dog also alerted on Kelly’s two sports utility vehicles, and the officers processed all three vehicles for forfeiture under Virginia state law.
B.
On March 7, 2007, a grand jury returned an indictment against Kelly and two co-defendants, both of whom subsequently pled guilty to various charges. Kelly was charged with conspiracy to distribute and possess with intent to distribute cocaine and cocaine base in violation of 21 U.S.C. § 846, two counts of possession with intent to distribute cocaine in violation of 21 U.S.C. § 841(a)(1) and 18 U.S.C. § 2, and one count of possession with intent to distribute ecstasy in violation of 21 U.S.C. § 841(a)(1) and 18 U.S.C. § 2.
Prior to trial, Kelly moved to suppress the evidence seized from his Lexus, claiming that the warrantless search of the vehicle violated the Fourth Amendment. Following a suppression hearing, the district court denied the motion. It held that the automobile exception to the warrant requirement applied and that the police had probable cause to search the vehicle, including its trunk, based on the totality of the circumstances. It stressed three factors supporting probable cause: (1) the presence of an individual that fit the description of Kelly’s cocaine supplier, (2) Kelly’s admission that there was cocaine in the vehicles, and (3) the drug detection dog’s positive alert on the Lexus.
Kelly pled not guilty, and a jury trial began on May 13, 2008. The government put on evidence that Kelly had been involved in a drug conspiracy since sometime around 2000. Several cooperating co-eonspirators testified that they personally purchased cocaine from Kelly, set up cocaine deals between Kelly and others, or otherwise witnessed Kelly distributing cocaine. Notably, Jose Jiminez explained that he brokered deals for cocaine and, on one occasion, ecstasy between Kelly and suppliers in New York City. Jiminez testified that he would travel to Virginia, stay at Kelly’s residence while Kelly distributed the drugs, and then transport some of the money back to New York City to pay the suppliers. The last of these drug deals, Jiminez testified, ended when the police arrested him and Kelly at Kelly’s residence. The government then presented witnesses and evidence regarding the seizure of drugs from the Lexus in 2006.
On May 16, 2008, the jury convicted Kelly of all counts. Subsequently, the district court imposed a sentence of life imprisonment, ten years supervised release, and a special assessment of $400.
On appeal, Kelly challenges the district court’s denial of his motion to suppress the evidence seized from the Lexus. He also appeals the district court’s denial of his motion for a mistrial based on prosecutorial misconduct and challenges the sufficiency of the evidence introduced at trial. We address each claim in turn and set forth additional facts as they become necessary.
II.
We first address the district court’s denial of Kelly’s motion to suppress. Kelly challenges both elements of the district court’s holding: that (1) the automobile exception to the warrant requirement applies here and (2) the police had probable cause to search the Lexus, including its trunk. We review the district court’s legal determinations
de novo
and its factual determinations for clear error. Because the district court denied Kelly’s motion, we construe the evidence in the light most favorable to the government.
United States v. Branch,
537 F.3d 328, 337 (4th Cir.2008).
We begin our analysis with a brief review of the law governing automobile searches. The Fourth Amendment generally requires the police to obtain a warrant before conducting a search. There is a well-established exception to this requirement, however, for automobile searches.
See, e.g., Carroll v. United States,
267 U.S. 132, 153, 45 S.Ct. 280, 69 L.Ed. 543 (1925). Under this exception, “[i]f a car is readily mobile and probable cause exists to believe it contains contraband, the Fourth Amendment thus permits police to search the vehicle without more.”
Pennsylvania v. Labron,
518 U.S. 938, 940, 116 S.Ct. 2485, 135 L.Ed.2d 1031 (1996) (per curiam). The scope of a search pursuant to this exception is as broad as a magistrate could authorize.
United States v. Ross,
456 U.S. 798, 825, 102 S.Ct. 2157,
72 L.Ed.2d 572 (1982). Thus, once police have probable cause, they may search “every part of the vehicle and its contents that may conceal the object of the search.”
Id.
There are two justifications for the automobile exception. The Supreme Court’s early cases were based on the mobility of the automobile. Unlike homes or other structures, cars “can be quickly moved out of the locality or jurisdiction in which the warrant must be sought.”
Carroll,
267 U.S. at 153, 45 S.Ct. 280. This rationale for the automobile exception is not one whose utility has diminished with time.
Carroll,
the seminal case, was decided in 1925, and the speeds at which automobiles are capable of traveling have only increased since that day.
See Scott v. Harris,
550 U.S. 372, 375, 127 S.Ct. 1769, 167 L.Ed.2d 686 (2007) (describing a car chase “at speeds exceeding 85 miles per hour”).
More recent cases provide a second justification for the exception. “Besides the element of mobility, less rigorous warrant requirements govern because the expectation of privacy with respect to one’s automobile is significantly less than that relating to one’s home or office.”
South Dakota v. Opperman,
428 U.S. 364, 367, 96 S.Ct. 3092, 49 L.Ed.2d 1000 (1976). It is true, of course, that cars harbor personal effects and that some vehicles, mobile homes for example, blend the properties of automobiles and residences. But, unlike a home, a ear has not been termed an owner’s castle and, again unlike a home, an automobile ventures out in public. This lesser expectation of privacy thus stems from the fact that cars as movable public objects are subject to “pervasive schemes of regulation.”
California v. Carney,
471 U.S. 386, 392, 105 S.Ct. 2066, 85 L.Ed.2d 406 (1985).
Kelly first contends that the automobile exception does not apply here because the police exercised control over the vehicle and had therefore eliminated any potential exigencies. As he points out, the officers at the scene were armed, had both Kelly and Jiminez handcuffed and under arrest, and possessed the keys to the Lexus. He further contends that he was placed in a police cruiser and that four to seven officers were present at the residence. Under these circumstances, he argues, there was not “even the slightest possibility that any potential evidence in the Lexus might be in danger of imminent destruction,” and the police should have obtained a warrant before conducting the search. Br. of Appellant at 13-14. In essence, Kelly urges us to recognize an exception to the automobile exception when the police have significantly reduced the likelihood that a car will be driven away.
As an initial matter, we note that officers may not have had as much control over this situation as Kelly suggests. They had no way of knowing whether Kelly’s girlfriend or one of his co-conspirators would arrive at the residence during the time it took them to secure a warrant, nor did they know whether there was a spare set of keys to the Lexus. Moreover, the Lexus was parked along a road where it was easily accessible to others.
But even if Kelly is correct that there was little risk that the Lexus would be driven away, it matters not. It is well established that the exception “does not have a separate exigency requirement” apart from the inherent mobility of the automobile.
Maryland v. Dyson,
527 U.S. 465, 467, 119 S.Ct. 2013, 144 L.Ed.2d 442 (1999) (per curiam). This is because “[e]ven in cases where an automobile [is] not immediately mobile, the lesser expectation of privacy resulting from its use as a
readily mobile vehicle justified] application of the vehicular exception.”
Carney,
471 U.S. at 391, 105 S.Ct. 2066. Thus, reviewing courts need not determine the probability in each case that someone would have driven the car away during the time it would have taken the police to secure a warrant.
Michigan v. Thomas,
458 U.S. 259, 261, 102 S.Ct. 3079, 73 L.Ed.2d 750 (1982) (per curiam). Instead, the exception applies as long as a car is “readily mobile” in the sense that it is “being used on the highways” or is “readily capable of such use” rather than, say, “elevated on blocks.”
Carney,
471 U.S. at 392-93, 394 n. 3, 105 S.Ct. 2066.
Following this precedent, we have previously declined to carve out exceptions to the automobile exception based on the degree of control police exercise over a vehicle. In
United States v. Brookins,
345 F.3d 231 (4th Cir.2003), for example, we applied the exception to the search of an unoccupied vehicle parked in a driveway even though the police could have blocked the vehicle from escaping.
Id.
at 237-38. Likewise, in
United States v. Gastiaburo,
16 F.3d 582 (4th Cir.1994), we applied the exception after the police had impounded a vehicle, rendering it “virtually impossible for anyone to drive the car away or tamper with its contents.”
Id.
at 586. As these cases recognize, “the justification to conduct a warrantless search under the automobile exception does not disappear merely because the car has been immobilized.”
Id.
These cases also reflect the well-recognized need to provide “clear and unequivocal guidelines to the law enforcement profession” in the context of automobile searches.
See California v. Acevedo,
500 U.S. 565, 577, 111 S.Ct. 1982, 114 L.Ed.2d 619 (1991) (internal quotations omitted). As this very case illustrates, whether the police exercise control over an automobile sufficient to eliminate any exigencies may turn on a number of imponderable factors. Were the applicability of the automobile exception to turn on such factors, it would be hard for courts to administer the exception in a coherent and predictable manner. The result would be a series of finely spun legal distinctions that would render the exception difficult for police to follow in quickly developing situations. Would the police, for example, have sufficient control over a vehicle solely because the driver was sitting in a police car? Would they have control if there were six or seven officers at the scene? Two or three? What about a passenger or two? Would the police have control if the driver had a flat tire by the roadside or if he parked his car in a driveway where the police could block his getaway? Would they have control if they possessed the keys to the ignition? And so on. Given the reduced privacy interests at stake, such doctrinal complexities are particularly undesirable in this context.
For these reasons, we reject Kelly’s attempt to create an exception to the automobile exception. Turning to the facts here, there is no doubt that Kelly’s Lexus was operational and therefore readily mobile. Consequently, the police were not required to obtain a warrant before searching the car or the containers therein.
C.
We now turn to whether the police had probable cause to conduct the warrantless search of the Lexus. Probable cause is “not readily, or even usefully, reduced to a neat set of legal rules.”
Illinois v. Gates,
462 U.S. 213, 232, 103 S.Ct. 2317, 76 L.Ed.2d 527 (1983). However, the Supreme Court has described it as “existing where the known facts and circumstances are sufficient to warrant a man of reasonable prudence in the belief that
contraband or evidence of a crime will be found.”
Ornelas v. United States,
517 U.S. 690, 696, 116 S.Ct. 1657, 134 L.Ed.2d 911 (1996). When assessing probable cause, we must examine the facts “from the standpoint of an objectively reasonable police officer,” giving “due weight to inferences drawn from those facts by ... local law enforcement officers.”
Id.
at 696, 699, 116 S.Ct. 1657.
Here, Kelly acknowledges that the police had probable cause based on the drug detection dog’s positive alert and for good reason.
See Florida v. Royer,
460 U.S. 491, 506, 103 S.Ct. 1319, 75 L.Ed.2d 229 (1983) (“[A] positive result [from a dog] would have resulted in ... probable cause”);
United States v. Jeffus,
22 F.3d 554, 557 (4th Cir.1994) (finding probable cause based on a positive alert);
see also United States v. Place,
462 U.S. 696, 707, 103 S.Ct. 2637, 77 L.Ed.2d 110 (1983);
United States v. Branch,
537 F.3d 328, 340 n. 2 (4th Cir.2008). But Kelly challenges the scope of the ensuing search. Although the dog’s alert at the driver’s door may have provided probable cause to search the passenger compartment, he argues, it did not provide probable cause to search the trunk, which was a few feet away from the exact spot of the alert. We disagree.
To be sure, probable cause to search one compartment or container within a car does not invariably provide probable cause to search the entire vehicle.
See California v. Acevedo,
500 U.S. 565, 580, 111 S.Ct. 1982, 114 L.Ed.2d 619 (1991). But it does not follow, as Kelly would have it, that a dog’s alert at one compartment cannot give probable cause to search another compartment simply because the latter is a few feet away. Unsurprisingly, Kelly cites no authority supporting such a bold proposition. The only case on which he relies notes instead that a “dog’s alerting [at a driver’s door] was sufficiently close to the trunk to give [the officer] probable cause to believe it contained contraband,”
United States v. Carter,
300 F.3d 415, 422 (4th Cir.2002). Kelly points us to no features of his car that would allow us to deviate from
Cartels
ultimate holding.
Probable cause is simply not so exacting a standard that it requires a dog to be able to pinpoint the location of drugs within a foot or two. Instead, it is a “commonsense” conception that deals with “the factual and practical considerations of everyday life.”
Ornelas,
517 U.S. at 695, 116 S.Ct. 1657 (internal quotation omitted). Dogs, of course, react not to the presence of drugs themselves but to their odors. And, as the K-9 officer in this case explained, odors travel within a car and seep through loose seals into the outside environment. Consequently, he continued, it is not uncommon for dogs to alert at the doors of vehicles where the seals tend to be heavily worn. Given these practical realities, we think it was reasonable to conclude that the odor which the dog detected may have travelled from the trunk, which is after all a logical place for drugs to be stored.
The district court alluded to several additional factors here. The first was the presence of an individual whom the police believed to be Kelly’s drug supplier. The police had just completed a year-long investigation in which they learned that Kelly’s main cocaine supplier was a Hispanic male from New York City. When they entered Kelly’s residence on the day of the search, they happened to find an individual meeting this description. What is more, the police knew that both Kelly and the Lexus had been absent from the residence the previous day. As the district court found, “it was logical to assume [Kelly] and his suspected cocaine source had arrived the night before in the dark green Lexus parked near [Kelly’s] residence.” Kelly questions the strength of this inference,
pointing out that the police did not have a detailed physical description of his supplier and had not observed Kelly trafficking drugs in the Lexus. But probable cause does “not require officials to possess an airtight case before taking action.”
Taylor v. Farmer,
13 F.3d 117, 121 (4th Cir.1993).
Second, Kelly nodded “yes” when asked if there was cocaine in the vehicles parked outside the residence. To be sure, he did not specify which of the three vehicles contained cocaine, but there was a fair probability that the drugs would be found in the Lexus, which had been driven the most recently.
There can thus be no question that the district court properly upheld the search of the Lexus for drugs, and the trunk and the backpack therein were logical places for the officers to look. Accordingly, we affirm the district court’s denial of Kelly’s motion to suppress.
III.
Next, Kelly challenges the district court’s denial of his motion for a mistrial based on prosecutorial misconduct. The relevant facts are as follows. The indictment in this case listed two prior occasions on which Kelly had been arrested for possessing drugs as overt acts in furtherance of the charged conspiracy. The first was a 2002 arrest in Norfolk for possession of a few grams of cocaine, and the second was a 2006 arrest in New York City for possession of marijuana. On the latter occasion, police discovered two hidden compartments in Kelly’s vehicle.
During her opening statement, the prosecutor referred to both of these arrests as evidence of Kelly’s longstanding involvement in the drug conspiracy. As the trial progressed, however, she only opted to present evidence regarding the second arrest. One of her witnesses explained that Kelly had not been prosecuted in connection with that arrest.
On three different occasions, Kelly moved for a mistrial based on these comments and evidence, claiming that the prosecutor was “essentially forcing [him] to take the stand” to explain the circumstances surrounding each arrest. Each time, the district court denied the motion. It specifically found that the prosecutor “[did not] force the defendant ... to have to take the stand to defend himself in any way.” In any event, it ruled, Kelly “[did not] suffer any type of prejudice” in light of the weight of the evidence against him. It also instructed the jury that Kelly was on trial only for the charges against him.
On appeal, Kelly argues that the prosecutor engaged in misconduct by referring to the two arrests and by presenting testimony regarding the latter one. This conduct, he contends, “forced [him] to make a Hobson’s choice whether to waive his absolute right [against] self-incrimination and testify so as to clarify the circumstances of his prior arrests, or to remain silent and have the jury draw impermissible conclusions based on those prior arrests.” Br. of Appellant at 22.
This claim is easily rejected. The prosecutor’s comments regarding the 2002 arrest addressed an overt act in the conspiracy and, in all events, were fleeting and in no way prejudiced Kelly in light of the overwhelming evidence against him.
As to his 2006 arrest, Kelly does not contend that evidence about this arrest was irrelevant or offered as a prior bad act “to show action in conformity therewith” as prohibited by Federal Rule of Evidence
404(b). Such claims would clearly be meritless: the government’s theory was that Kelly routinely travelled to New York City to obtain drugs, and it offered evidence of this arrest to show that he was in that very city driving a car with hidden compartments for stowing contraband. Instead, Kelly argues that this evidence violated his Fifth Amendment right against compelled self-incrimination by putting him in an undesirable predicament.
This claim too is meritless. Kelly claims that the government was forcing him to explain that he had not been convicted in connection with the 2006 arrest, but any such pressure was relieved once a government witness testified explicitly that Kelly had not even been prosecuted. Even assuming that the government’s evidence did somehow put pressure on Kelly to take the stand (which he never actually did), his constitutional rights were not infringed. Evidence by its nature builds pressure to rebut it — that’s what the adversary system is about. “That the defendant faces ... a dilemma demanding a choice between complete silence and presenting a defense has never been thought an invasion of the privilege against compelled self-incrimination.”
Williams v. Florida,
399 U.S. 78, 84, 90 S.Ct. 1893, 26 L.Ed.2d 446 (1970). It is often true that “there are undoubted pressures — generated by the strength of the government’s case against him — pushing the criminal defendant to testify,” but such pressures do not “constitute ‘compulsion’ for Fifth Amendment purposes.”
Ohio Adult Parole Auth. v. Woodard,
523 U.S. 272, 287, 118 S.Ct. 1244, 140 L.Ed.2d 387 (1998). Thus, we agree with the district court’s conclusion that the government did not improperly pressure Kelly to take the stand. Accordingly, we affirm the court’s denial of Kelly’s motion for a mistrial.
IV.
Lastly, Kelly claims that the evidence at trial was insufficient to support any of his four convictions. His primary argument is that the majority of the government’s witnesses were unworthy of belief because they were testifying pursuant to plea agreements. But it is well established that “determinations of credibility are within the sole province of the jury and are not susceptible to judicial review.”
United States v. Burgos,
94 F.3d 849, 863 (4th Cir.1996) (internal quotation omitted). Here, each of the government’s cooperating witnesses disclosed whether he was testifying pursuant to a plea agreement. It was up to the jury to weigh each witness’s credibility in light of such agreements, and it is not our role to second-guess these determinations.
After careful review, we find that Kelly’s other arguments are likewise meritless and conclude that the evidence presented was sufficient to support the jury’s verdict.
V.
For the foregoing reasons, the judgment of the district court is
AFFIRMED.