UNITED STATES v. PEREDA AKA AKA AKA AKA AKA AKA (2026)

OPINION

One morning, police officers found Binisio Nicolas Pereda—a state early-release supervisee subject to warrantless search conditions—in a trailer located near his girlfriend's parents’ home. Officers then searched the trailer and a nearby pickup truck thought to be used by Pereda. They found ammunition in the trailer and drugs in the truck. We must decide whether officers had probable cause to believe that Pereda (1) resided in or controlled the trailer, and (2) owned or controlled the pickup truck. Because the officers had probable cause for both searches, the searches were consistent with the Fourth Amendment. We thus reverse the district court's order granting Pereda's motion to suppress and remand for further proceedings.

I.

Background

In 2016, Pereda was sentenced to a 14-year term of imprisonment in state custody for several violent felonies. In 2022, he was released from prison early and placed in post-release community supervision. As part of his release, he agreed to several supervision conditions, including a GPS monitoring condition and a warrantless search condition. The search condition provided that Pereda, his “residence,” or “any other property under [his] control may be searched without a warrant day or night by an agency of the supervising county, any peace officer, or any enforcement officer.” The conditions also required Pereda to inform his supervising officers of his “residence” and to report any “changes or anticipated changes in residence ․ in advance.” He also could not travel more than 50 miles from his “residence” without prior written approval.

In March 2024, Santa Barbara County Probation Officer Jose Macedo took over Pereda's supervision. Based on GPS monitoring, Officer Macedo noticed that Pereda would often leave his residence at late hours to visit Los Angeles at odd times. Officer Macedo suspected that Pereda was “involved with some sort of drug activity.” So Pereda's supervision team decided to perform a compliance check on June 27, 2024.

Officers first needed to determine Pereda's residence. Pereda reported to supervising officers that his parents’ home in Lompoc, California was his residence. But Officer Macedo observed that Pereda consistently stayed at an address in Goleta, California, which was believed to be Pereda's girlfriend's parents’ house. According to Pereda's GPS location, he had consistently pinged at the Goleta address from March to June 2024—including overnight. The GPS data placed Pereda “outside” of the main house at the Goleta address, slightly to its left:

At the same time, Officer Macedo observed that it was “not uncommon” for Pereda to stay at the Lompoc address “for hours at a time” or visit various parts of Lompoc.

Before the scheduled compliance check, Officer Macedo reviewed Pereda's GPS location for the prior two or three days. That data placed Pereda by the Goleta address. Officer Macedo also submitted reports placing Pereda by the Goleta address for the eight days leading up to the check, usually just next to the main house. Officer Macedo also knew that his predecessor had conducted home visits for Pereda at the Goleta, not Lompoc, address. And Pereda's former supervising officer had admonished Pereda for failing to update his reported residence from the Lompoc address to the Goleta address.

Based on this information, the supervision team conducted the compliance check at the Goleta address. When officers arrived, they knocked on the front door of the residence, and Pereda's girlfriend's mother answered. When the officers asked for Pereda, she replied that Pereda was “in the trailer” and that the officers “would find him” there. The officers then noticed a trailer roughly 15 feet from the house, explaining why the GPS data kept placing Pereda slightly outside the main structure. The trailer was behind a fence, propped up on poles and cinderblocks, and connected to the house with an extension cord. Officers also noticed a white Ford F-150 truck parked closely next to the trailer.

Officers then knocked on the trailer door and asked for Pereda. Pereda's girlfriend, without opening the door, responded that Pereda was in the house, not the trailer— contradicting what her mother had told the officers. The officers continued to knock, and Officer Macedo heard a male voice from inside the trailer. So he called out to Pereda by name and ordered him to exit. Pereda complied, and an officer performed a pat-down search.

Around this time, Santa Barbara County Sherriff's Office Detective Amjadi arrived on scene. When he arrived, Detective Amjadi recognized a white Ford F-150 pickup truck parked next to the trailer. Detective Amjadi had seen Pereda driving a white Ford F-150 pickup truck around five times over the previous several weeks—including on the same street where the truck was parked on the day of the search. On one of those five occasions, Detective Amjadi and his partner ran the truck's license plate number, confirming that Pereda was the registered owner.

After officers ordered Pereda's girlfriend to exit the trailer, Detective Amjadi and Officer Macedo began to search it, where they found ammunition and glass pipes. During the search, Officer Macedo told Detective Amjadi that he wanted to search the pickup truck parked next to the trailer. Based on his prior observations, Detective Amjadi replied that he thought the white Ford F-150 was the one Pereda had been driving. Without checking the license plate or registration, the officers recovered the keys from the trailer and searched the truck. Inside the truck's cab, officers found large quantities of methamphetamine, fentanyl, and heroin.

The officers then arrested Pereda. A federal grand jury indicted Pereda with possession with intent to distribute controlled substances and with being a felon in possession of ammunition. See 21 U.S.C. § 841(a)(1); 18 U.S.C. § 922(g)(1).

Pereda moved to suppress the drugs, drug paraphernalia, and ammunition as fruits of an unreasonable search under the Fourth Amendment. The district court granted Pereda's suppression motion. The district court concluded that the officers lacked probable cause to believe (1) that the trailer was Pereda's residence, and (2) that the truck was property under Pereda's control. The district court then excluded all the evidence. The government appealed. See 18 U.S.C. § 3731 (permitting “[a]n appeal by the United States ․ from a decision or order of a district court suppressing or excluding evidence”).

We “review the validity of a warrantless search de novo” and the district court's underlying factual findings “for clear error.” United States v. Franklin, 603 F.3d 652, 655 (9th Cir. 2010) (simplified).

II.

Warrantless Searches Valid Under The Parole Exception

The Fourth Amendment to the United States Constitution provides that “[t]he right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated[.]” U.S. Const. amend. IV. Warrantless searches are generally “per se unreasonable.” United States v. Estrella, 69 F.4th 958, 964 (9th Cir. 2023) (simplified). But there are exceptions to this rule. United States v. Brown, 996 F.3d 998, 1004 (9th Cir. 2021). One of the recognized exceptions to the warrant requirement is the so-called parolee exception. Estrella, 69 F.4th at 964. Under the parolee exception, “[a] search of a parolee that complies with the terms of a valid search condition will usually be deemed reasonable under the Fourth Amendment.” United States v. Cervantes, 859 F.3d 1175, 1183 (9th Cir. 2017).

A.

The Trailer

1.

The Trailer as a Residence

Pereda's post-release community supervision search conditions authorized a search of his “residence.” When a search condition permits a warrantless search of a parolee's “residence,” “law enforcement officers must have probable cause to believe that the parolee is a resident of the house to be searched.” United States v. Grandberry, 730 F.3d 968, 973 (9th Cir. 2013) (simplified). The probable cause needed to search a parolee's residence is the same well-established probable-cause standard used throughout criminal law. It “is not a high bar[.]” Kaley v. United States, 571 U.S. 320, 338, 134 S.Ct. 1090, 188 L.Ed.2d 46 (2014). Instead, it “is a fluid concept that is not readily, or even usefully, reduced to a neat set of legal rules[.]” District of Columbia v. Wesby, 583 U.S. 48, 57, 138 S.Ct. 577, 199 L.Ed.2d 453 (2018) (simplified). “Probable cause as to residence exists if an officer of reasonable caution would believe, based on the totality of the circumstances, that the parolee lives at a particular residence.” United States v. Barry, 140 F.4th 1105, 1109 (9th Cir. 2025) (simplified); see also Kaley, 571 U.S. at 338, 134 S.Ct. 1090 (explaining that probable cause “requires only the kind of fair probability on which reasonable and prudent people, not legal technicians, act.” (simplified)). Because probable cause is a totality of the circumstances standard, a court errs when it “view[s] each fact in isolation,” rather than “consider[ing] the whole picture” that the facts create. Wesby, 583 U.S. at 60–61, 138 S.Ct. 577 (simplified). In other words, courts can't use a “divide-and-conquer analysis” to determine whether probable cause existed. Id. at 61, 138 S.Ct. 577.

Considering the whole picture, the officers had ample probable cause to believe that the trailer was Pereda's residence.

First, before the search, officers knew that Pereda was likely not living at his reported address in Lompoc, California. Instead, for roughly three months, officers believed that Pereda had consistently stayed at the Goleta address.

Second, they also knew that Pereda's former probation officer conducted home visits at the Goleta address and admonished him for not living in Lompoc.

Third, GPS data placed him just outside the Goleta address over a prolonged period. GPS data showed that Pereda's location clustered around the exact spot that officers found the trailer for at least eight days before the search, including overnight.  1 Officer Macedo, who had reviewed Pereda's GPS data on a nearly daily basis, confirmed the same pattern for the “two or three days” before the search.

Fourth, the officers’ observation on the day of the search reinforced their belief that Pereda lived in the trailer. They first knocked on the door of the house at the Goleta address and asked for Pereda. The resident, Pereda's girlfriend's mother, told them that they would find Pereda in the trailer. When they approached the trailer, they saw that it had the markings of a residence. For example, an extension cord connected the trailer to the house. And while the trailer could be moved after being hitched to a truck, it was immobile and propped up on stilts and cinderblocks at the time of the search. Finally, a white Ford F-150 pickup truck was parked right next to the trailer—the same type of vehicle Pereda had registered in his name and in which he had been seen driving around five times before the search.

Fifth, the officers’ interactions with Pereda the morning of the search validated their beliefs that he resided in the trailer. After knocking on the trailer door, Pereda's girlfriend falsely stated that he was not there, suggesting that she knew he was not supposed to be living there. When she opened the door, she was only partially clothed, indicating that they were both sleeping there. Finally, the officers heard Pereda's voice from inside the trailer.

Adding all these facts up, there was a “fair probability” that the trailer was Pereda's residence. See Kaley, 571 U.S. at 338, 134 S.Ct. 1090. Under the totality of the circumstances, the officers “reasonabl[y] infer[red],” United States v. Hamilton, 131 F.4th 1087, 1095 (9th Cir. 2025), that Pereda was residing in the trailer. And that's all they needed to conduct the search. So the warrantless search of the trailer was a valid search of Pereda's “residence” under his supervised release conditions.

Rather than consider the “whole picture,” the district court engaged in the type of “divide-and-conquer” analysis that the probable-cause inquiry prohibits. See Wesby, 583 U.S. at 60–61, 138 S.Ct. 577. Take the district court's consideration of the GPS data. The district court first discounted the GPS data for the two to three days before the search because “the officers had no way of knowing what Pereda was doing at the pinged locations in this short period of time—whether he was smoking a cigarette, walking the dog, talking to a neighbor, or even staying in one of the detached buildings at the back of the Goleta property.” The district court then isolated the data alone and concluded that it “was not sufficient evidence” to show that Pereda was residing in the trailer.

But this was the wrong approach to considering the evidence. First, “an officer is not required to eliminate all innocent explanations for a suspicious set of facts[.]” Pennsylvania v. Dunlap, 555 U.S. 964, 964, 129 S.Ct. 448, 172 L.Ed.2d 321 (2008) (Roberts, C.J., dissenting from denial of certiorari). Second, courts don't consider the relative strength of each fact on its own; that's because “the whole is often greater than the sum of its parts[.]” Wesby, 583 U.S. at 60–61, 138 S.Ct. 577. The officers didn't present the GPS data as sufficient on its own—the data was one part of the whole picture used to reach probable cause. And while other inferences may be drawn from the GPS data, under the totality of the circumstances, it was reasonable to infer that Pereda's sustained presence on the precise location of the trailer—including overnight—meant that he lived in the trailer. While it's true that officers didn't have metaphysical certainty of Pereda's residence, that is not what probable cause requires. They needed far less.

Nor was it appropriate to reduce the probable-cause analysis into a box-checking inquiry. For example, the district court relied heavily on the fact that officers did not observe Pereda using a key to the trailer or recover a set of keys to the trailer from him. The district court also faulted the officers for not running a registration check on the trailer before searching it. While the presence of a key or a registration check would be helpful and have been part of past searches we've approved, see Grandberry, 730 F.3d at 976, it's important to recognize that “the ultimate question whether probable cause exists is fact-intensive, and cannot be answered by cross-checking a list of factors,” id. (simplified). Whether certain factors are met does not answer the probable-cause question; instead, the inquiry is always case-specific and cannot be “reduced to a neat set of legal rules[.]” Wesby, 583 U.S. at 57, 138 S.Ct. 577.

Our dissenting colleague relies on United States v. Howard, 447 F.3d 1257 (9th Cir. 2006), to defend the district court. Dissent at –––– – ––––. To be sure, probable cause “turns on the assessment of probabilities in particular factual contexts[.]” Wesby, 583 U.S. at 64, 138 S.Ct. 577 (simplified); see also Dissent at ––––. But contrary to the dissent's contention, there is no two-tiered probable cause standard—one for arresting suspects and a higher one for searching a parolee's residence. Even in the parolee search context, probable cause remains a totality-of-the-circumstances standard. Barry, 140 F.4th at 1109.

Howard synthesized our past cases applying the parolee exception. Howard, 447 F.3d at 1265–67. “In considering the cases where this Court upheld the search of an address not reported by a parolee,” we explained, “certain patterns clearly emerge.” Id. at 1265. Those “patterns” are as follows: (1) “the parolee did not appear to be residing at any address other than the one searched,” (2) “the officers had directly observed something that gave them good reason to suspect that the parolee was using his unreported residence as his home base,” (3) “the parolee had a key to the residence in question,” and (4) “either the parolee's co-resident or the parolee himself identified the residence in question as that of the parolee.” Id. at 1265–66.

But these “patterns” are just that—patterns. We have recognized that although they may “guide[ ] our application of the probable cause standard,” “the ultimate question whether probable cause exists is fact-intensive, and cannot be answered by cross-checking a list of factors.” Grandberry, 730 F.3d at 976 (simplified). Howard’s patterns are not factors that must be present for probable cause to exist, and we have not treated them that way. See Barry, 140 F.4th at 1110 (making fact-intensive probable cause determination without relying on Howard); Franklin, 603 F.3d at 655–57 (same). After all, treating Howard as announcing binding factors would flout the Supreme Court's repeated explanations that probable cause “is a fluid concept” that cannot “be reduced to a neat set of legal rules[.]” Wesby, 583 U.S. at 57, 138 S.Ct. 577. So although Howard can offer useful guidance, the probable cause determination remains a fact-intensive one.

Applying that fact-intensive standard, we conclude that there was a fair probability that Pereda was residing in the trailer. So the officers had probable cause to search the trailer according to Pereda's parole conditions. The search was thus consistent with the Fourth Amendment.

2.

The Trailer As Property Under Control

The trailer search was valid for another reason. Pereda's post-release community supervision terms expressly permitted a warrantless search of “property under [his] control.” Under that provision, officers may search a parolee's property if the parolee “exhibits a sufficiently strong connection to the property in question to demonstrate control over it.” United States v. Korte, 918 F.3d 750, 754 (9th Cir. 2019) (simplified). To search a vehicle according to a parolee's “property under control” provision, officers “must have probable cause to believe that the supervisee owns or controls the vehicle to be searched.” United States v. Dixon, 984 F.3d 814, 822 (9th Cir. 2020). The officers had probable cause to believe that Pereda owned or controlled the trailer.

First, the trailer is subject to the “property under control” provision because the trailer is a vehicle. To be sure, the trailer was propped up on stilts and cinderblocks. But it was still a travel trailer with rear wheels that could be hitched to the back of a truck. And, under California law, a “travel trailer, truck camper, or camping trailer,” even one that, as here, is “without motive power,” is a “[r]ecreational vehicle.” Cal. Health & Safety Code § 18010(a) (emphasis added). So for California, the trailer is a vehicle. That the trailer was not moving or hitched to a motor vehicle is immaterial. After all, the trailer was still “readily capable of such use” and “obviously readily mobile.” See California v. Carney, 471 U.S. 386, 392–93, 105 S.Ct. 2066, 85 L.Ed.2d 406 (1985). With just a few minutes of work, the trailer could be barreling down a highway, crossing city, county, or state lines.

Second, Pereda showed a “sufficiently strong connection” to the trailer to demonstrate his “control over it.” Korte, 918 F.3d at 754 (simplified). In Korte, we concluded that “[o]fficers observ[ing]” the defendant “putting things inside the trunk” of his vehicle—even once—“illustrated a sufficiently close nexus to the trunk itself.” Id. at 755. The officers here had much stronger evidence than the officers in Korte. Rather than watching Pereda put something in the trailer one time, officers had three months of data placing Pereda at the trailer. They recognized his voice from inside when they approached the trailer, and they watched him exit the trailer with his partially clothed girlfriend. Pereda's conduct between March and June, and his conduct on the day of the compliance check, “illustrated a sufficiently close nexus to the [trailer].” Id. Based on this strong connection, the officers had “probable cause to believe that” Pereda controlled “the [trailer] to be searched.” Dixon, 984 F.3d at 822.

And nothing in Grandberry precludes applying the “property under control” provision to trailers. There, officers tried to search a residential apartment unit under both the “residence” and “property under control” provisions of the parolee's supervised release. Grandberry, 730 F.3d at 980. Based on our precedent, we ruled that the “property under control” provision cannot apply to “residential locations.” Id. at 980–81. We thought that extending the “property under control” provision to residences would render the “residence” provision a “nullity.” Id. at 982. But Grandberry explicitly recognized that the “property under control” provision applies to “searches ․ of vehicles[.]” Id. at 981 (compiling cases). Indeed, Grandberry cited approvingly People v. Boyd, 224 Cal. App. 3d 736, 740, 750–51, 274 Cal.Rptr. 100 (Cal. App. 1990), which involved a parole search of a travel trailer that the Boyd court considered “clearly” valid. See Grandberry, 730 F.3d at 981 n.11; see also Boyd, 224 Cal. App. 3d at 744, 274 Cal.Rptr. 100. And it discussed the residence provision in the context of a residential apartment building. Id. at 976–78. It said nothing about a situation where, as here, the parolee appears to be residing in a vehicle.

So Grandberry teaches that the “property under control” provision cannot apply to an apartment—but it can apply to a vehicle. The trailer here is both a residence and a vehicle. Grandberry thus does not foreclose a search according to the “property under control” provision.

Because officers had probable cause that Pereda owned or controlled the trailer, the trailer search can also be justified according to the “property under control” provision in Pereda's post-release search conditions.

B.

The Ford F-150

The officers also reasonably searched the truck parked next to the trailer. As noted above, Pereda's post-release community supervision terms permitted a warrantless search of any “property under [Pereda's] control.” That provision applies to the search of a parolee's vehicle. See id. at 981. So Pereda's supervised release conditions permitted a search of his vehicle if the officers had probable cause to believe that he owned or controlled the vehicle searched. Dixon, 984 F.3d at 822. Once again, probable cause is “not a high bar” and “requires only the kind of fair probability on which reasonable and prudent people, not legal technicians, act.” Kaley, 571 U.S. at 338, 134 S.Ct. 1090 (simplified).

That standard is met here. Detective Amjadi had seen Pereda driving a white, single cab, Ford F-150 pickup truck about five times, including on the same street where the truck was parked on the day of the search. He also knew that Pereda had a Ford F-150 registered to him. And when Detective Amjadi arrived at the scene, he saw Pereda come out of the trailer, and saw a white, single cab, Ford F-150 parked directly in front of the trailer's door. Indeed, the truck's main cabin was mere feet away from the trailer's main door. Simply, the truck was parked in a place where someone who had driven to the trailer probably would have parked it:

Based on his prior observations, the truck's proximity to the trailer, and watching Pereda exit the trailer, Detective Amjadi made a “reasonable inference[ ]” that Pereda owned or controlled the truck. United States v. Struckman, 603 F.3d 731, 740 (9th Cir. 2010). And these “commonsense judgments and inferences” support probable cause. Hamilton, 131 F.4th at 1095 (simplified).

The district court again focused on what the officers didn't do. It noted that the officers didn't see Pereda in the truck around the time of the search, that they didn't ask whose truck it was, and that they didn't run a registration check on the truck before searching it. No doubt, these additional steps would have further supported probable cause. But their absence is not fatal. The relevant question isn't whether additional facts would have bolstered the officers’ conclusion. Instead, the question is whether “the facts and circumstances known” at the time, id. at 1094 (simplified), create a “fair probability” that the parolee owned or controlled the vehicle, Kaley, 571 U.S. at 338, 134 S.Ct. 1090. That the officers could have taken additional steps before searching the truck does not change that the probable-cause standard was met here.

Based on the totality of the circumstances, the facts available to the officers established a fair probability that Pereda owned or controlled the white pickup truck. Armed with probable cause, the officers validly searched the truck according to Pereda's supervised release terms.

III.

Conclusion

For these reasons, the searches of the trailer and the pickup truck were consistent with the Fourth Amendment. We thus reverse the grant of the motion to suppress and remand for further proceedings consistent with this opinion.

REVERSED and REMANDED.

Binisio Nicolas Pereda agreed to a number of parole conditions upon his early release from state prison in 2022. Of relevance here, Pereda agreed to warrantless searches of “[him], [his] residence, and any other property under [his] control.” Thus, if the location or property to be searched was Pereda's “residence” or “property under [his] control,” the search would be proper. If not, the search would be unlawful. We apply a “relatively stringent standard” to determine whether officers, prior to searching, were reasonably sure they were at the right house, United States v. Grandberry, 730 F.3d 968, 974, 976 (9th Cir. 2013), and reasonably sure they had the right person's property, United States v. Dixon, 984 F.3d 814, 822 (9th Cir. 2020). This standard is protective of third parties’ Fourth Amendment rights. It recognizes that merely inviting a parolee to one's home or parking next to a parolee's vehicle should not subject innocent third parties to intrusive searches. 1

I respectfully dissent because the majority ignores this standard that is recognized in our case law. Applying it here, I would affirm the district court's determination that the officers lacked probable cause to believe that the trailer was Pereda's residence and that the truck was Pereda's property. I address the search of the trailer and truck in turn.

A. Search of the Trailer as Pereda's Residence

1. The District Court's Faithful Application of Our Case Law

The district court correctly considered the totality of the circumstances, including the “factors” described in Grandberry, to determine that the officers lacked probable cause to believe that the trailer was Pereda's residence. These factors were first outlined in United States v. Howard, 447 F.3d 1257, 1265 (9th Cir. 2006). I agree with the majority that satisfaction of all the Howard factors is not a prerequisite for a district court to find probable cause; they are not a checklist. Grandberry, 730 F.3d at 976. But our precedent establishes that these factors “guide[ ] our application of the probable cause standard.”  2 Id. The majority's assertion that the probable-cause-as-to-residence standard “is the same well-established probable-cause standard used throughout criminal law” ignores the context of the warrantless searches at issue. Majority at ––––. The majority's casual approach is not supported by our case law.

The guiding factors we employ to analyze probable cause in this context are prompted by California's parole search conditions. Unlike in other suppression cases, if the parole search conditions are not met, the search is not lawful. The cases relied upon by the majority are removed from this context. They deal with a grand jury's finding of probable cause to indict, Kaley v. United States, 571 U.S. 320, 338, 134 S.Ct. 1090, 188 L.Ed.2d 46 (2014), as well as a police officer's finding of probable cause to arrest, D.C. v. Wesby, 583 U.S. 48, 56, 138 S.Ct. 577, 199 L.Ed.2d 453 (2018), United States v. Hamilton, 131 F.4th 1087, 1096 (9th Cir. 2025), Pennsylvania v. Dunlap, 555 U.S. 964, 964, 129 S.Ct. 448, 172 L.Ed.2d 321 (2008) (Roberts, C.J., dissenting from denial of certiorari). Whether someone committed a crime, however, is a distinct inquiry from whether someone is living at a particular residence. Both ask what “an officer of reasonable caution would believe, based on the totality of the circumstances,” Grandberry, 730 F.3d at 975 (citation modified), but they involve different relevant factors.

It is therefore troubling that the majority acknowledges, but does not analyze, the Howard factors, apparently determining that they constitute “the type of ‘divide-and-conquer’ analysis that the probable-cause inquiry prohibits.” Majority at ––––. Not so. A district court may examine the totality of the circumstances by viewing multiple relevant factors “cumulatively.” Grandberry, 730 F.3d at 976. Here, these factors include:

(1) “the parolee did not appear to be residing at any address other than the one searched”; (2) “the officers had directly observed something that gave them good reason to suspect that the parolee was using his unreported residence as his home base”; (3) “the parolee had a key to the residence in question”; and (4) “either the parolee's co-resident or the parolee himself identified the residence in question as that of the parolee.”

Id. (citation omitted). Howard held that these factors establish a “relatively stringent standard,” as recognized in United States v. Franklin, 603 F.3d 652, 656 (9th Cir. 2010), and in Grandberry, 730 F.3d at 976. Howard, 447 F.3d at 1262. After analyzing our existing cases, Howard concluded that those finding probable cause as to residence rely on “strong evidence” and “very strong facts.” Id. at 1263–64, 1267. Further, “[w]hen presented with weaker facts,” Howard acknowledged that “we have not hesitated to rule that officers could not justify a search for lack of probable cause.” Id. at 1265. Grandberry reiterated that “ ‘[t]here must be strong evidence’ that the parolee resides at the address.” Grandberry, 730 F.3d at 976.

Respectfully, the majority's assertion that we have previously ignored Howard when assessing probable cause as to residence is not reflected in our case law. Majority at ––––. In Franklin, we recognized that Howard established a “relatively stringent standard” in cases with parolees thought to be living in the homes of third parties, though we found that standard inapplicable in cases with parolees thought to be living in motel rooms that they had personally rented. Franklin, 603 F.3d at 656–57. In Barry, we listed the Howard factors, as identified in Grandberry, and determined that probable cause existed when an officer observed the searched residence in-person to corroborate a tip that the parolee was selling drugs from there, the parolee identified the residence as his own, and the parolee had the key to the residence on his person; i.e., when at least three Howard factors weighed in favor of finding probable cause. United States v. Barry, 140 F.4th 1105, 1109–10 (9th Cir. 2025).

Examining the first factor here, there was clearly evidence known to the officers at the time of the search that Pereda was “residing at any address other than the [trailer].” See Howard, 447 F.3d at 1265. The district court determined that the officers “had reliable evidence that Pereda was residing both in Lompoc and Goleta.” For example, while Officer Macedo determined from GPS tracks that Pereda “consistently stayed” at the Goleta address, he also stated that it was “not uncommon for Pereda to stay at the Lompoc address for ‘hours at a time’ or visit ‘various locations’ throughout Lompoc.” Further, as recognized by the district court, while some of the facts support an inference that Pereda was residing at the Goleta address, they do not demonstrate that Pereda was residing at a trailer at a different address, separated by a tall, solid fence. In other words, evidence that sufficiently supports probable cause as to one residence (the Goleta address) cannot independently support probable cause as to an entirely different residence (the trailer). Here, the officers did not even know that the trailer existed before the search. Because there was some evidence that Pereda was residing at two locations other than the trailer, the officers’ “paltry effort” to determine where Pereda resided “weighs against” a probable cause finding. Grandberry, 730 F.3d at 978 n.9.

The evidence of this factor in this case is a far cry from other cases in which we have affirmed a probable cause finding. See United States v. Dally, 606 F.2d 861, 862–63 (9th Cir. 1979) (explaining that officer could not find defendant at his reported address and defendant failed to return a message left for him there); United States v. Watts, 67 F.3d 790, 795 (9th Cir. 1995), rev'd on other grounds, 519 U.S. 148, 117 S.Ct. 633, 136 L.Ed.2d 554 (1997) (reasoning that after more than fifty visits over fourteen months, officer found defendant at his reported address once, but it “lacked the usual signs of residency”); United States v. Conway, 122 F.3d 841, 842–43 (9th Cir. 1997) (noting that officer visited defendant's reported address twenty-one times and found him there once, and learned that his only belongings there were “a pair of socks”). In this case, the officers conducted no in-person surveillance of the reported address in Lompoc or Pereda's suspected address in Goleta.

Turning to the second factor, the district court correctly concluded that there was insufficient evidence of Pereda using the trailer as his “home base.” Even with the GPS data, “the officers had no way of knowing what Pereda was doing at the pinged locations in this short period of time—whether he was smoking a cigarette, walking the dog, talking to a neighbor, or even staying in one of the detached buildings at the back of the Goleta property.” Unlike in Dally, Watts, Harper, Conway, and Barry, the officers did not observe Pereda, for example, (1) enter or exit the trailer, (2) drive to and from the trailer, (3) take out the trash near the trailer, (4) carry his dry cleaning near the trailer, (5) chat with his neighbors near the trailer, (6) park overnight near the trailer, (7) change his clothes in the trailer, (8) receive mail at the trailer, or (9) permit “known associates” to park at the trailer. See Dally, 606 F.2d at 862; Watts, 67 F.3d at 793; United States v. Harper, 928 F.2d 894, 896 (9th Cir. 1991), rev'd on other grounds by United States v. King, 687 F.3d 1189 (9th Cir. 2012); Conway, 122 F.3d at 843; Barry, 140 F.4th at 1108. And unlike in Watts and Barry, the officers were not following up on tips that Pereda resided at or sold drugs from the trailer. See Watts, 67 F.3d at 795; Barry, 140 F.4th at 1107.

As for the third factor, the district court appropriately noted that the officers “did not observe Pereda using a key to the trailer or recover a set of keys to the trailer from his person.” As discussed in Dally, Harper, Conway, and Barry, this weighs against a finding of probable cause, though it is not dispositive. See Dally, 606 F.2d at 862; Harper, 928 F.2d at 895; Conway, 122 F.3d at 843; Barry, 140 F.4th at 1110; Grandberry, 730 F.3d at 979. After all, one would expect someone who was living in a residence to at least have the keys to that residence.

Fourth, as the district court recognized, “[n]o one identified the trailer as Pereda's residence prior to the search.” This is distinguishable from Conway, 122 F.3d at 843, and Watts, 67 F.3d at 793. This is also different from Barry, where the officers informed the defendant that they would search his apartment, and the defendant did not “express any surprise or deny living there.” Barry, 140 F.4th at 1110.

Viewing these factors “cumulatively,” the officers arrived at the Goleta address with reliable evidence that Pereda resided at the Lompoc address and the Goleta address. They did not know that the trailer existed until Pereda's girlfriend's mother told them he was “in” it. They saw an electrical extension cord extending from the trailer to the Goleta house, but had no information whether it was Pereda's residence, his girlfriend's residence, or even his girlfriend's parents’ residence. As the record reflects, the officers had not observed any actions by Pereda at the trailer that were probative of residence. They did not hear or ask anyone if Pereda lived there or observe or ask anyone about a key to the trailer. Pereda's overnight stays from the past two or three days at an area that appeared to be near the Goleta house in the GPS data “is the only fact [considered so far that is] consistent with [Pereda's] living there, and it is a weak indication, on its own.” See Grandberry, 730 F.3d at 979. When combined with Pereda's girlfriend's mother's comment and the sound of Pereda's voice behind his partially unclothed girlfriend, the indicia that the trailer was Pereda's residence remain insufficient under our case law. Rather, as the district court concluded, the cumulative facts known to the officers show only Pereda's status as an overnight guest who “spent the night [in the trailer] occasionally.” See Howard, 447 F.3d at 1262.

2. The Majority's Other Errors

The majority reaches a different conclusion by ignoring the Howard factors and relying on facts that the district court did not find. Absent clear error, we cannot cast aside the district court's factual findings. See Franklin, 603 F.3d at 655 (“The district court's factual finding[s] [in a motion to suppress] are reviewed for clear error.”).

First, the majority ignores the district court's finding that the officers “had reliable evidence that Pereda was residing both in Lompoc and Goleta.” Instead, the majority makes its own, contrary factual finding: that “before the search, officers knew that Pereda was likely not living at his reported address in Lompoc.” Majority at ––––. Unless the majority concludes that the district court committed clear error, it cannot simply rewrite the factual findings on this point, which is especially probative of the first Howard factor.

Similarly, the majority ignores the district court's finding that the officers were not aware of the trailer and did not know why Pereda's GPS location pings in the days leading up to the search were outside of the Goleta house. The district court made forceful findings on this point:

The trailer was not simply an extension of the Goleta residence that the officers had not yet seen—the trailer was parked in a completely different area alongside at least one vehicle and adjacent to a tall, solid fence on the other side of the Goleta residence. It was akin to a neighboring home or apartment with a separate street address and entrance.

Instead, the majority states that the officers “belie[ved] that Pereda lived in the trailer” based on the GPS data alone, a belief that was “reinforced” when they were told he was in the trailer by Pereda's girlfriend's mother and heard his voice. Majority at ––––.

Finally, the majority relies on GPS data collected from “at least eight days before the search, including overnight” that was “submitted [in] reports” by Officer Macedo. Majority at ––––, ––––. Tellingly, the majority cites to Officer Macedo's testimony on direct examination for this assertion. The district court, however, acted within its discretion to credit Officer Macedo's testimony on cross-examination, in which he admitted that he only reviewed “the last two or three days [of GPS data]” prior to the compliance check. The maps referred to by the majority, which include eight days of GPS data, were created a month before the evidentiary hearing in this case. They were not created or consulted by the officers before the search was conducted in June 2024 and therefore cannot inform the probable cause determination. The data underlying these maps, described by the majority in a footnote, does not alter the analysis.

The majority defends its reliance on this data as an effort to “explain what ‘consistently stayed’ means.” Majority at –––– n.1. In the majority's view, when the district court found that “[Officer] Macedo determined that Pereda ‘consistently stayed’ [at the Goleta address]” based on GPS tracking, it meant to find that Officer Macedo knew Pereda had stayed “just outside the Goleta address ․ for at least eight days before the search, including overnight.” Majority at ––––. The district court, however, never made this finding. Instead, the district court found that Officer Macedo “admitted that he only reviewed GPS data for the two to three days prior to the compliance check,” and that the officers “had no way of knowing what Pereda was doing at the pinged locations in this short period of time.” The majority claims it does not disturb this finding, but proceeds to weigh the evidence independently and conclude that Officer Macedo knew of Pereda's GPS locations for the eight days prior to the search, rather than just two to three days. It is not our role to weigh the evidence. 3 See McClure v. Thompson, 323 F.3d 1233, 1240 (9th Cir. 2003).

Further, the majority attempts to rely on the officers’ day-of observations of Pereda and what they believed to be his truck, and the fact that Pereda's girlfriend's mother told the officers he was “in” the trailer. But as the district court recognized, we cannot draw an inference that the trailer was Pereda's residence “simply because [he] happens to be seen” in the residence. Franklin, 603 F.3d at 657. The majority's reliance on Pereda's girlfriend's lack of clothing and the officers’ sight of Pereda when he exited the trailer, which may have “indicat[ed] that they were both sleeping there,” is similarly infirm. See Watts v. County of Sacramento, 256 F.3d 886, 890 (9th Cir. 2001) (“[T]he mere fact that Pryor answered the door of his girlfriend's home in his boxer shorts did not establish a reasonable belief that he lived there.”).

The majority's critiques of the district court are therefore unwarranted. The district court did not “divide and conquer” by “isolat[ing]” the GPS data. Majority at ––––. Rather, as discussed above, the district court properly focused on the GPS data because the totality of the evidence relied on by the government only showed that Pereda was physically present in the trailer the day of the search, which is not sufficient. Furthermore, the district court's twenty-four page opinion was not a “box-checking” exercise: it assessed relevant factors cumulatively to consider the totality of the circumstances. Majority at ––––. Finally, the majority states that the district court “relied heavily on the fact that officers did not observe Pereda using a key to the trailer or recover a set of keys to the trailer from him.” Majority at ––––. Not so. The district court explicitly noted that the lack of keys was “not dispositive on its own”—a faithful application of Grandberry, 730 F.3d at 979.

For these reasons, I would affirm the district court's well-reasoned ruling and hold that the officers lacked probable cause to believe that the trailer was Pereda's residence.

B. Search of the Trailer as Property Under Pereda's Control

After deeming the search of the trailer lawful as a search of Pereda's residence, the majority concludes that, alternatively, the search was lawful because the officers had probable cause to believe it was property under Pereda's control. This conclusion is unnecessary and, more importantly, directly at odds with the rule set forth in Grandberry.

In Grandberry, the government made a similar argument: that even if the officers could not search an apartment as the defendant's “residence,” they could search it as “property under [his] control.” Grandberry, 730 F.3d at 980. We described this argument as a “loophole” and rejected it, citing our precedent, the text of the parole conditions, and the doctrinal underpinnings of probable cause. Id. at 981–82. We should have done the same here.

The majority attempts to skirt this issue by cherry picking quotations from Grandberry. Notably missing from the majority opinion is the following discussion from Grandberry: “As we are bound by our precedent, we conclude that the ‘property under your control’ provision cannot refer to a place where someone else, but not the parolee, lives.” Id. at 980–81. Grandberry further explained that where “it is abundantly clear that the searched property is a residence,” the search is only justified under these parole conditions “when the officers have probable cause that the parolee lives at a residence.” Id. at 982.

The majority acknowledges that the trailer here was a residence. Majority at ––––. Under Grandberry, that should be the end of the inquiry: because the trailer is a residence, it cannot also be “property under [a person's] control.” Such an interpretation would “render the separate ‘your residence’ term a nullity.” Grandberry, 730 F.3d at 982. Furthermore, it would “eviscerate the privacy interests of non-parolees in the home, where Fourth Amendment rights are sacrosanct.” Id. (citation modified). Plainly, the government cannot fall back on the “property under your control” provision to search a residence when it fails to establish probable cause as to residence.

Respectfully, the majority's reliance on Grandberry’s mention of “vehicles” misconstrues dicta. The passage from Grandberry states:

[T]he government has cited no case—and we have found none—applying the “property under your control” search condition to a residence. Instead, the fairly large number of cases applying that term uniformly involve searches either of vehicles or of items found in a residence.

Grandberry, 730 F.3d at 981. The court subsequently cited People v. Boyd, 222 Cal.App.3d 541, 271 Cal. Rptr. 738, 750–51 (1990), with the following explanatory parenthetical: “search of handbag inside a trailer.” Id. at 981. I agree with the district court that this “parallel phrasing would suggest that Grandberry likened a trailer to a residence, not a vehicle.” Nonetheless, the accompanying footnote in Grandberry makes clear that “[t]here was no challenge to the entry into the trailer [in Boyd], and so no occasion to consider whether that entry was in fact authorized[.]” Grandberry, 730 F.3d at 981 n.11. The majority's assertion that Grandberry approvingly cited Boyd, which “involved a parole search of a travel trailer,” is therefore misleading. Majority at –––– – ––––.

While the Grandberry court did not have occasion to consider whether a trailer should be treated as a residence or a vehicle in this context, it set forth the following rule: in order to search “where, as here, it is abundantly clear that the searched property is a residence,” there must be probable cause to believe that it is the parolee's residence. Grandberry, 730 F.3d at 982. Applying this rule, the facts known to the officers at the time of the search made it abundantly clear that the trailer was a residence: it was propped up with supporting stakes, connected to the Goleta house through an extension cord, Pereda's girlfriend's mother told them it was occupied, and Pereda's girlfriend exited the trailer without pants. It therefore could not be searched under this provision unless there was probable cause to believe it was Pereda's residence, which the officers failed to establish.

C. Search of the Truck as Property Under Pereda's Control

Turning to the search of the truck, the majority once again omits the relevant standard. In Dixon, we discussed the reasons behind our probable-cause-as-to-residence standard, which “protects the interest of third parties.” 984 F.3d at 822 (citation modified). We then determined that there was “no reason to depart from this standard with respect to a supervisee's vehicle.” Id. We expressly rejected a lower “reasonable suspicion” standard because it would “run[ ] the risk of officers conducting intrusive searches on vehicles that have no connection to the individual subject to the search condition.” Id.

Under this standard, the district court appropriately determined that, at best, the officers only had reasonable suspicion to support the search of the truck. The majority relies on Officer Amjadi's prior sightings of Pereda in a white Ford F-150, which he knew was registered to Pereda, as well as the fact that a white Ford F-150 was parked right next to the trailer in which Pereda was located. These facts, viewed cumulatively, do not establish probable cause that Pereda owned or controlled the searched truck.

The white Ford F-150 is, “year after year,” the “best-selling vehicle in America.”  4 Indeed, a similarly colored pick-up truck was located on the other side of the trailer in this very parking lot:

The ubiquitousness of this type of vehicle makes it clear that any prior sightings of Pereda were insufficient to establish probable cause. With so many white Ford F-150s on the road, how could the officers have been reasonably sure that the previously observed truck and the searched truck were the same truck? Plainly, they were not. Because Pereda did not “exhibit a sufficiently strong connection” to the searched truck to “demonstrate control over it,” Dixon, 984 F.3d at 818 (citation modified), the officers only suspected that the previously observed truck and the searched truck were the same.

True, the searched truck was close to the trailer, and Pereda was in the trailer. But Pereda's girlfriend was also in the trailer. And “[s]everal other cars were parked nearby.” Because there were numerous vehicles on this street, which could have belonged to Pereda, Pereda's girlfriend, or any of their neighbors or their guests, the truck's proximity to the trailer, even combined with the prior sightings of Pereda, was insufficient to establish probable cause. The officers’ actions jeopardized the privacy rights of all of the third parties with cars parked on this street, which is exactly what our case law aims to avoid. Dixon, 984 F.3d at 822.

The majority faults the district court for “focus[ing] on what the officers didn't do.” Majority at ––––. The district court should not be faulted for faithfully applying our case law. In comparing the factual similarities and differences between the instant case and Dixon, as well as United States v. Dodd, No. 24-CR-078-JHC, 2025 WL 1360554, at *11 (W.D. Wash. May 9, 2025), the district court correctly noted the absence of relevant factors here: the officers did not see Pereda in or near the truck, did not recover keys to the truck from him, 5 did not ask any questions about it, and did not verify the license plate or run a registration check. Accordingly, I agree with the district court that the officers failed to “have a sufficient basis to believe [that Pereda] owned or controlled” the truck before they searched it. Dixon, 984 F.3d at 818–19.

***

Because neither the search of the trailer nor the search of the truck were supported by probable cause, I would affirm the district court. I respectfully dissent because the majority ignores the analytical approach to California parolee cases that we have adopted in our case law and degrades the Fourth Amendment protections it affords to non-parolees.

It is worth noting that our case law does not ask for much. The government would not be here today if any officer on the scene had simply asked Pereda, his girlfriend, or his girlfriend's parents “who lives here,” “whose truck is this,” or “who has the keys to the trailer or truck?” Even “I am going to search your trailer,” or “I am going to search your truck,” directed to Pereda without any denial from him, might have been enough. See Barry, 140 F.4th at 1108. The officers could have engaged in any in-person surveillance of Pereda using the trailer as his residence prior to the day of the search. Similarly, in mere minutes before the search, they could have examined the permit in the trailer's window or ran a registration check on the trailer or truck. I agree with the district court that “[t]he officers’ failure to do so reflects a rush to judgment, not reasoned investigation,” warranting suppression.

FOOTNOTES

1.   Our dissenting colleague notes that the maps showing eight days of GPS data “were created a month before the evidentiary hearing” and “were not created or consulted by the officers before the search.” Dissent at ––––. But the data in those reports were available before the search, and Officer Macedo—who participated in the compliance check—had monitored the GPS data for roughly three months. See Franklin, 603 F.3d at 656 (“Probable cause requires that the facts available to the officer would warrant a man of reasonable caution in the belief that [the trailer] was [Pereda's] residence at the time.” (simplified)). And as the district court found, Officer Macedo “credibly testified that, based on location monitoring, he observed that Pereda ‘consistently stayed’ at the Goleta address.” Pereda does not challenge this finding. Contrary to the dissent's suggestion, we do not disturb the district court's finding that Officer Macedo reviewed two or three days of GPS data before the search. Dissent at –––– – ––––. We merely look to the record on appeal and the GPS data available to both Officer Macedo and the district court to explain what “consistently stayed” means. Cf. United States v. W.R. Grace, 504 F.3d 745, 766 (9th Cir. 2007) (explaining that we consider the record that was before the district court).

1.   Our concern for third parties is well-demonstrated by the instant case. Pereda's girlfriend was forced to exit the trailer and walk down the street in her underwear, while her belongings in the trailer were indiscriminately searched.

2.   The majority uses the term “patterns” instead of “factors,” quoting Howard, 447 F.3d at 1265. Majority at ––––. But in Grandberry, we described these patterns as “factors,” reflecting our holding that “Howard’s synthesis of our precedents ․ guides our application of the probable cause standard to the facts known to the [o]fficers before they conducted the warrantless search[.]” Grandberry, 730 F.3d at 976. We continued to use the term “factors” in Barry. United States v. Barry, 140 F.4th 1105, 1109 (9th Cir. 2025). Thus, even assuming this difference in terminology is meaningful, the majority errs in eliding Grandberry.

3.   I fail to see how the cited portion of United States v. W.R. Grace, 504 F.3d 745, 766 (9th Cir. 2007), supports the majority's handiwork. Majority at –––– n.1.

4.   See Sean Tucker, The Most Popular Car Colors Are, Once Again, Neutral, KBB, (Dec. 15, 2020), https://www.kbb.com/car-news/the-most-popular-car-colors-are-once-again-neutral/ [https://perma.cc/Q9TS-NBUV].

5.   Rather, the keys to the truck were recovered from Pereda's girlfriend's purse, casting even greater doubt on whether the officers were reasonably sure that Pereda, not his girlfriend, owned or controlled the searched truck.

BUMATAY, Circuit Judge:

Opinion by Judge Bumatay; Dissent by Judge Paez