PER CURIAM Opinion; Opinion by Judge FERGUSON: Opinion by Judge KOZINSKI; Concurrence by Judge TASHIMA; Concurrence by Judge RONALD M. GOULD; Concurrence by Judge PAEZ.
PER CURIAM.A majority of the en banc court (Judge Ferguson, joined by Chief Judge Schroeder and Judges Pregerson, Tashima, Paez and Berzon) concludes that the officers did not have probable cause to enter Johnson’s *898property, and were not in hot pursuit when they searched the area outside the mushroom shed. That same majority also concludes that whether the search took place within the curtilage is a question that must be determined in the first instance by the district court.
A different majority (Judge Kozinski, joined by Judges Trott, T.G. Nelson, Silverman, Gould and Paez) concludes that, under Ornelas v. United States, 517 U.S. 690, 116 S.Ct. 1657, 134 L.Ed.2d 911 (1996), the determination of the curtilage must be reviewed de novo on appeal. Of those, five judges (Judge Kozinski, joined by Judges Trott, T.G. Nelson, Silverman and Gould) conclude that the search took place outside the curtilage.
On the probable cause issue, the decision of the district court is REVERSED. On the curtilage issue, we REMAND to the district court for a determination of whether the search lay within the curtilage and any other issues necessary to the disposition of this case. The panel retains jurisdiction over the case in the event of any further appeals.
FERGUSON, Circuit Judge, with whom Chief Judge SCHROEDER and Circuit Judge PREGERSON join; Circuit Judge TASHIMA joins with respect to all of the opinion except Part III.B a, b, c, and d; Circuit Judge PAEZ joins with respect to all of the opinion except Part III.A; and Circuit Judge BERZON joins with respect to all of the opinion except for Part III.B and the final two sentences of Part III.A:
I.
In an attempt to apprehend another person who was a misdemeanor suspect last seen 30 minutes previously and whose whereabouts were unknown, state sheriff officers broke into Michael Johnson’s fenced and locked residential yard on his rural Washington property without a warrant. While searching his yard, officers smelled marijuana in a detached shed. As a result of this warrantless search, a search warrant was issued and Johnson was subsequently indicted on one count of manufacturing marijuana in violation of 21 U.S.C. §§ 841(a)(1) and (b)(1)(B). Prior to entering a conditional plea of guilty, Johnson filed a motion to suppress the evidence gained as result of the search.
The district court denied Johnson’s motion, determining that the search was justified under the hot pursuit and exigent circumstances exceptions to the warrant requirement of the Fourth Amendment. In affirming the district court, the three-judge panel of this Court assumed that the shed was not in an open field but was part of the curtilage. United States v. Johnson, 207 F.3d 538, 543-44 (9th Cir.2000). However, the district court did not make any finding or conclusions page regarding the open field-curtilage question.
We therefore:
1. REMAND the case to the district court for factual findings and conclusions on whether the shed was in an open field or part of the curtilage, a matter not developed by the district court.
2. REVERSE the district court on the issue of the warrant requirement, a question that was fully developed in the district court and by the three-judge panel.
II.
In February 1998, Washington State Child Protective Service (“CPS”) asked Deputy Chris Kading of the Skagit County Sheriffs Office to check on the welfare of the children living in Steven Dustin Smith’s residence in rural Skagit County, Washington. Deputy Kading ran a check on Smith and discovered that Smith had *899five outstanding arrest warrants for misdemeanor offenses.1
On February 14, 1998, Deputy Fading drove by Smith’s residence and noticed that he was standing outside his home. As Deputy Fading pulled into the driveway, Smith began to run. Deputy Fading ordered Smith to stop, which he did. Once he confirmed Smith’s identity, Deputy Fading told him that he was under arrest and ordered him to turn and place his hands behind his back. According to Deputy Fading, Smith became very agitated. He began clenching his fists and looking at the deputy’s gun. Deputy Fading unsuccessfully attempted to calm Smith with “verbal judo.” Smith made no sudden moves towards Fading, but instead turned to look off down the road. When he turned back to Fading, Fading sprayed his face with pepper spray because, he felt, the “verbal judo” was not working. Smith dropped to his knees and covered his eyes with his hands. Deputy Fading grabbed Smith and attempted to handcuff him, but Smith broke free and began running down the highway.
The attempted arrest took place at a trailer park located at 3090 South Skagit Highway. Approximately 157 feet west down the highway, a steep and rocky driveway led to the home of Smith’s mother and step-father. Approximately 1254 feet in the same direction lay the driveway of Defendant Michael Johnson. All of these locations were south of the highway.
Deputy Fading returned to his vehicle and began pursuing Smith. As he drove down the highway, he saw Smith leave the road and run south straight into the woods. At this point, Smith was approximately halfway to Defendant Michael Johnson’s driveway. Deputy Fading quickly lost sight of him in the thick brush. Although his testimony is somewhat confusing, Fading apparently believed that Smith would remain in the woods between Johnson’s residence and Smith’s mother’s house.2
Fading radioed for back up and requested a canine unit. He turned down a long, twisting driveway and followed it back 250 yards into the woods until he came to a 14-foot locked gate with a “No Trespassing” sign. A 5-foot high cyclone fence extended in both directions from the gate. This was Johnson’s driveway. Fading then returned to the Smith residence for his pepper spray, came back, and waited by the locked gate for backup.
Approximately 15 minutes later, Deputy John Rose was directed by Fading to drive *900to Smith’s mother’s house and wait there. Ten minutes later, Deputy Sigman met Kading at the gated entrance to Johnson’s yard. The three deputies discussed the situation and formulated a plan in which Kading and Sigman would enter Johnson’s property and proceed through the woods toward Smith’s mother’s house.
In order to pursue Smith, the officers felt they needed to get through the locked gate. They could not open the padlock; instead, according to Kading, there was enough “play” in the locking mechanism to allow the officers to “manipulate” the hasp of the gate and gain entry while keeping the padlock locked. Once inside the gate, the officers drove up the driveway approximately 50 yards and parked their vehicle in a yard. On the right side of the driveway was Johnson’s house. On the left side of the driveway, approximately 30 yards from the house, was a large dog kennel enclosed by a chain link fence. Behind the dog kennel was a small “mushroom shed.”3 The shed was about 40-50 yards from the house. None of the residential area had been visible from the gate.
The first thing the officers did was knock on the door of Johnson’s home to see if he was in and alert him to their presence. When they received no answer, they began searching his property for Smith. First, they checked a covered patio near the back of the residence to see if Smith was hiding there. Next, they checked the outer area of the kennel. The officers then peered into two old vehicles parked next to the kennel and looked under a blue tarp that was spread out near the cars. Finally, the officers followed a trail behind the kennel to the mushroom shed, which was locked with a new padlock. When the officers were within one or two feet of the door of the mushroom shed, Kading smelled what he knew from training was marijuana. He tried to open the door to the shed, but the padlock prevented him. At this point, the deputies left Johnson’s property and drove to Smith’s mother’s house. They did not continue their planned search in the woods, and Smith was never found.
A state search warrant was subsequently obtained based on Kading’s and Sig-man’s observations while on Johnson’s property. On February 19, 1998, the warrant was executed and 553 marijuana plants were recovered. On April 23, 1998, Johnson was federally indicted on one count of manufacturing marijuana in violation of 21 U.S.C. §§ 841(a)(1), and 841(b)(1)(B).
Prior to entering a conditional plea of guilty, Johnson filed a motion to suppress, arguing that the deputies violated the Fourth Amendment when they conducted a warrantless search of his property. The district court judge held an evidentiary hearing and denied the motion, basing her ruling primarily on the existence of “hot pursuit” or “exigent circumstances.” Johnson subsequently pled guilty to the filed charges, but preserved his ability to appeal the denial of his motion to suppress.
III.
Although both the parties and the district court focused attention on the “exigent circumstances” and “hot pursuit” exceptions to the Fourth Amendment warrant requirement, it also must be determined whether the fenced area around the residence that was searched was subject to Fourth Amendment protection.
*901A.
The Fourth Amendment protection against warrantless searches extends to the curtilage around one’s home. “Whether an area is within the protected curtilage of a home is an essentially factual inquiry” that we review for clear error. United States v. Soliz, 129 F.3d 499, 502 (9th Cir.1997). The Second Circuit has “assumed, without deciding,” that a review of a curtilage determination was affected by Supreme Court’s holding in Ornelas v. United States, 517 U.S. 690, 699, 116 S.Ct. 1657, 134 L.Ed.2d 911 (1996), that determinations of probable cause and reasonable suspicion should be subject to de novo review. United States v. Reilly, 91 F.3d 331, 331 (2d Cir.1996). On the other hand, the Seventh Circuit has explicitly cited Ornelas for the proposition that curtilage is a factual determination to be reviewed under a “clearly erroneous” standard. United States v. Shanks, 97 F.3d 977, 979 (7th Cir.1996).
In this case, the facts regarding the curtilage issue and any determination based therein are very complex, as demonstrated by the discussions in this opinion and the dissent. Regardless of what type of review must ultimately be made, it is vital for the district court to first make findings of fact upon which that review can be based. Because a majority of the judges on the panel reach the dispositive issue and agree that the search violated Fourth Amendment protections, we leave for another panel the question of resolving what standard of review this Court applies to curtilage determinations. See United States v. Crespo de Llano, 830 F.2d 1532, 1542 n. 2 (9th Cir.1987) (holding that disagreement over standard of review was dictum where a majority agreed on the ultimate conclusion). The Court should wait for a case in which the district court makes explicit findings about curtilage before determining our standard of review for this issue. This case is therefore remanded to the district court for a determination of the curtilage issue. See United States v. Furrow, 229 F.3d 805, 817 (9th Cir.2000); United States v. Traynor, 990 F.2d 1153, 1156-57 (9th Cir.1993).
In United States v. Dunn, 480 U.S. 294, 301, 107 S.Ct. 1134, 94 L.Ed.2d 326 (1987), the Supreme Court directed that questions of curtilage should be
resolved with particular reference to four factors: the proximity of the area claimed to be curtilage to the home, whether the area is included in an enclosure surrounding the home, the nature of the uses to which the area is put, the steps taken by the resident to protect the area from observation by people passing by.
The Court stressed that these factors cannot be “mechanically applied,” but are merely “useful analytical tools” to determine whether an area is to be protected from unconstitutional searches and seizures. Id. The position taken by the dissent illustrates why this Court should not determine curtilage questions with no guidance from the district court as the factfinder. The findings by the dissent, as demonstrated below, miss essential elements in the determination of curtilage.
B.
In undertaking the four-factor analysis outlined in Dunn, the district court must be mindful of two additional factors that are important here. First, the government has conceded that it has the burden of proof in this case to show that the search was not within the curtilage of Johnson’s residence. Second, the Johnson property had been under surveillance by members of the Skagit County Inter-Local Drug Enforcement Unit, who suspected *902that Johnson was growing marijuana on his property. At no point during their investigation did these officers enter the area inside Johnson’s gated yard, which he contends is the curtilage of his rural property.
a. Proximity
Generally, “there is not any fixed distance at which curtilage ends.” United States v. Depew, 8 F.3d 1424, 1427 (9th Cir.1993). It must be determined on a case-by-case basis. Dunn, 480 U.S. at 301, 107 S.Ct. 1134. The Second and Third Circuits have noted the importance of considering whether the area in question is in a rural, urban, or suburban setting. See United States v. Reilly, 76 F.3d 1271, 1277, on reh’g, 91 F.3d 331 (2d Cir.1996) (concerning a rural property); United States v. Acosta, 965 F.2d 1248, 1256 (3d Cir.1992) (concerning an urban property); see also United States v. Seidel, 794 F.Supp. 1098, 1103 (S.D.Fl.1992) (concerning a suburban property). These courts have reasoned that the curtilage of a home in a rural area could extend farther than the curtilage of a home in an urban or suburban setting. Reilly, 76 F.3d at 1277 (stating that “on a large parcel of land, a pond 300 feet away from a dwelling may be as intimately connected to the residence as is the backyard grill of the bloke next door”).
The dissent does not distinguish between a yard and a field in a rural area. Calling a fenced-in area around a rural home an “open field” does not make it so. It is still a yard, even if somewhat larger than yards in urban areas. The realities of rural country life dictate that distances between outbuildings will be greater than on urban or suburban properties and yet still encompass activities intimately associated with the home; this is the nature of the “farmstead.” Recently this Court, in United States v. Furrow, 229 F.3d 805, 817 (9th Cir.2000), endorsed this analysis in adopting a district court’s finding that 100 feet was within the curtilage based in part on the “rural nature of the premises.”
b. Enclosure
The second factor analyzes whether the area is included within an enclosure surrounding the home. “[F]or most homes, the boundaries of the curtilage will be clearly marked; and the conception defining curtilage — as the area around the home to which the activity of home life extends — is a familiar one easily understood from our daily experience.” Dunn, 480 U.S. at 302, 107 S.Ct. 1134 (quoting Oliver v. United States, 466 U.S. 170, 182, n. 12, 104 S.Ct. 1735, 80 L.Ed.2d 214 (1984)). While not conclusive, “fencing configurations are important factors in determining curtilage.” Dunn, 480 U.S. at 301 n. 4, 107 S.Ct. 1134. In rural pieces of property, as here, natural boundaries such as thick trees or shrubberies may also indicate an area “to which the activity of home life extends.” Dunn, 480 U.S. at 302, 107 S.Ct. 1134 (quoting Oliver, 466 U.S. at 182, 104 S.Ct. 1735). See, e.g., Reilly, 76 F.3d at 1277 (finding that hedgerows and thick trees created a sufficient enclosure to determine curtilage).
Johnson’s property consists of over 12 acres of wooded land. Within the twelve-plus acres, a relatively small yard is enclosed by a five-foot high fence. This section encloses the house, the dog kennel, and the mushroom shed. Although there are some smaller fences within this yard, none of the internal fencing segregates the house from the shed. As this yard is surrounded by dense woods and underbrush, the only access to the mushroom shed is from Johnson’s driveway and residence.
*903c. Use
Prior to beginning their search, the officers possessed no objective data that the mushroom shed was not used for intimate activities associated with the home. Other circuits that have addressed the issue have found that officers must have “objective data” about the use of the area prior to entry. See Daughenbaugh v. City of Tiffin, 150 F.3d 594, 599 (6th Cir.1998) (finding search unreasonable when officers lacked prior objective knowledge that area searched was not used for activities of the home); Reilly, 76 F.3d at 1279 (2nd Cir.1996) (rejecting the government’s argument that the officers had “objective data” about the use of the outbuilding because the officers smelled marijuana after they entered the property); United States v. Swepston, 987 F.2d 1510, 1515 (10th Cir.1993) (same).
In Dunn, although the Supreme Court relied on information obtained both before (aerial photographs) and after (smell of phenylacetic acid) the search began, it emphasized that it found “especially significant” the fact that the law enforcement officials possessed “objective data” that the barn in question was used to manufacture drugs before entering the property. Dunn, 480 U.S. at 302, 107 S.Ct. 1134. Justice Scalia’s concurrence joined in the majority opinion with the exception of this paragraph. Id. at 305, 107 S.Ct. 1134 (Scalia, J., concurring). In his opinion, the only significant factor was that the barn was being used for illegal activity; it did not matter whether the officers knew it prior to their entry. Id. Dunn requires that when determining the “use” of an area, the officers cannot rely, as was done in this case, exclusively on information they learn after the search begins.
We have never held that an officer lacking any prior objective knowledge of the use of an outbuilding may approach it free of Fourth Amendment constraints. See, e.g., Depew, 8 F.3d at 1426-1427; Calabrese, 825 F.2d at 1350.
d. Visibility
The fourth Dunn factor focuses on the steps taken by Johnson to prevent observation of the area from passers-by. No part of the residential area is visible from the highway or from the gate. Johnson purchased the property precisely because it was in a rural and secluded area, and one of his reasons for constructing the fence was to keep out neighbors. Only two other people (the meter man and propane man) had keys to the padlock. Under these circumstances, Johnson appears to have made every effort to prevent the public from observing the residential portion of his property. See Depew, 8 F.3d at 1428 (outbuilding within curtilage where Depew made sincere effort to keep public from viewing his property).
The dissent suggests that the fact that Johnson’s fences did not block visibility is important to this inquiry. The uncontro-verted testimony of the officers, however, indicated that the woods and underbrush surrounding Johnson’s yard did block visibility. Indeed, this is why Smith was so difficult to follow. Physical boundaries may protect one’s privacy. See Reilly, 76 F.3d at 1277; Daughenbaugh, 150 F.3d at 599; Schroeder, 129 F.3d at 442. The dissent’s position, taken to its logical extreme, would place a considerably greater economic burden on rural residents who wish to fence in their property and find that chain-link fence is the most cost-effective way to do so. Privacy interests, however, do not depend on the wealth of a suspect.
State law is also relevant to determining the reasonableness of police activities un*904der the Fourth Amendment. Reed v. Hoy, 909 F.2d 324, 330 n. 5 (9th Cir.1989) (citing Tennessee v. Garner, 471 U.S. 1, 15-16, 105 S.Ct. 1694, 85 L.Ed.2d 1 (1985)).4 Washington courts have consistently held that its state constitution and laws are more protective of privacy than federal laws:
[T]he presence of the long history of territorial and state laws prohibiting trespass indicates that Washington places important emphasis on a person’s right to exclude others from his or her private property, regardless of the size or developed state of that property.
State v. Johnson, 75 Wash.App. 692, 879 P.2d 984, 990 (1994). The officers who intruded upon Johnson’s property were local and state officials to whom the Washington Constitution applies. The state constitution, however, creates a heightened expectation of privacy from intrusion by these officers.
Under the Washington Constitution, the test is whether the government intruded upon the defendant’s “private affairs.” State v. Myrick, 102 Wash.2d 506, 688 P.2d 151, 154-55 (1984). See supra note 7. In State v. Thorson, 98 Wash.App. 528, 990 P.2d 446, 448-50 (1999), the Washington Court of Appeals was asked to consider the legitimacy of a search where officers on a rural island strayed from the property for which they had a search warrant to another parcel. In holding that the search violated the Washington Constitution’s privacy protection, the court focused on subjective expectations of privacy and stated that
Myrick requires us to look to the nature of the property, the expectation of privacy it reasonably supports, and the nature of the intrusion to answer the ultimate question: Whether the government’s intrusion violated a privacy interest which citizens of this state have traditionally and justifiably held safe from governmental trespass absent a warrant.
Id. at 449. One of the officers who conducted the search in this case testified as to the kennel, located near the shed, that “if I would have been a private citizen driving up and parked where I did, I’d believe that the kennel would have been attached to that residence.” Despite the fact that local drug enforcement had suspected “for some time” that Johnson was growing marijuana on his property — and apparently had performed fly-overs of the area — they did not feel at liberty to enter and investigate the yard around his house.
Based on the combination of the (1) the rural setting, (2) the fence around the home and shed, (3) the lack of objective data pointing to illegal activity prior to entry, and (4) the inability to see the shed from the “open fields,” one could find that the shed was “so intimately tied to the home itself that it should be placed under the home’s ‘umbrella’ of Fourth Amendment protection.”. Dunn, 480 U.S. at 301, 107 S.Ct. 1134.
IV.
While the issue of curtilage in this ease remains unsettled, the district court fully developed its ruling that the warrant-less search of Johnson’s property was justified and not entitled to Fourth Amendment protections. Our inquiry now is whether the officers’ warrantless search of *905Johnson’s property is justified under the Constitution’s Fourth Amendment protections. If it is not, the marijuana plants seized as result of that search should be suppressed as fruit of the poisonous tree. Wong Sun v. United States, 371 U.S. 471, 487, 88 S.Ct. 407, 9 L.Ed.2d 441 (1963). We review de novo the validity of a war-rantless search. United States v. Van Poyck, 77 F.3d 285, 290 (9th Cir.1996).
“It is a ‘basic principle of Fourth Amendment law’ that searches and seizures inside a home without a warrant are presumptively unreasonable.” Payton v. New York, 445 U.S. 573, 586, 100 S.Ct. 1371, 63 L.Ed.2d 639 (1980). The presumption of unreasonableness can be overcome, however, when the police confront an exigent circumstance like a fleeing felon. See Welsh v. Wisconsin, 466 U.S. 740, 749-50, 104 S.Ct. 2091, 80 L.Ed.2d 732 (1984). In these situations, the exigent circumstance reheves the police of the obligation of obtaining a warrant. Murdock v. Stout, 54 F.3d 1437, 1441 (9th Cir.1995). The exigent circumstance does not, however, reheve the pohce of the need to have probable cause for the search. Id. As a result, when the government rehes on the exigent circumstances exception, it still must satisfy two requirements: first, the government must prove that the officer had probable cause to search the house; and second, the government must prove that exigent circumstances justified the warrantless intrusion. United States v. Lai, 944 F.2d 1434, 1441 (9th Cir.1991). Based on the record before us, the government failed to satisfy either requirement.
A.
Probable cause is hardly a new concept. See U.S. CONST. amend. IV (“The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated, and no Warrants shall issue, but upon probable cause ...” ). For over 75 years, the Supreme Court has stated that probable cause exists when the “facts and circumstances” before the officer are sufficient to warrant a person of reasonable caution to believe that the items sought will be found in the place to be searched. See Brinegar v. United States, 338 U.S. 160, 175-76, 69 S.Ct. 1302, 93 L.Ed. 1879 (1949); Carroll v. United States, 267 U.S. 132, 162, 45 S.Ct. 280, 69 L.Ed. 543 (1925). The Supreme Court has emphasized that probable cause “demands” factual “specificity” and “must be judged according to an objective standard.” Terry, 392 U.S. at 21-22 n. 18, 88 S.Ct. 1868. “Anything less would invite intrusions upon constitutionally guaranteed rights based on nothing more than inarticulate hunches, a result this Court has consistently refused to sanction.” Id. at 22, 88 S.Ct. 1868. (emphasis added).
By his own testimony, Officer Hading was led to Johnson’s property by no more than a “gut feeling” that Smith would be there. The Supreme Court, however, has made it clear that “hunches” are insufficient to establish reasonable suspicion, let alone probable cause. See Illinois v. Wardlow, 528 U.S. 119, 123-24, 120 S.Ct. 673, 145 L.Ed.2d 570 (2000) (stating that reasonable suspicion, which is “less demanding than probable cause,” requires an officer to articulate more than an “unparticularized suspicion or ‘hunch’ ”); Terry v. Ohio, 392 U.S. 1, 27, 88 S.Ct. 1868, 20 L.Ed.2d 889 (1968) (same); United States v. Kerr, 817 F.2d 1384, 1387 (9th Cir.1987) (emphasizing that even for a Terry stop, “hunches alone will not withstand constitutional scrutiny”).
Officer Hading did not provide a single objective fact to support his hunch or establish that it provided probable cause to rummage around Johnson’s yard. At the *906suppression hearing, Officer Kading testified that he initially thought that it was “logical” that Smith would turn left (southeast) in the woods and head directly toward his parent’s house. He was so confident of this prediction that when he radioed for backup, he told the responding officers to proceed to Smith’s parent’s house. Soon afterward, however, he had second thoughts. He stated that he tried to get inside of Smith’s head and outsmart him. He speculated that if Smith thought that the officers would go to his parent’s house, Smith would do the opposite and run toward Johnson’s house (southwest). When asked how he came to this conclusion, Officer Kading admitted that it was just a “gut feeling.”
We find no objective facts in the record to support Officer Kading’s “gut feeling.” For example, Officer Kading does not claim that he saw footprints leading toward Johnson’s property, heard sounds coming from Johnson’s property, or observed broken branches on the trees leading toward Johnson’s property. The only thing that Officer Kading knew was that a half hour earlier, Smith had run into the woods halfway between Johnson’s driveway and Smith’s driveway and that Smith’s mother lived nearby, in the opposite direction of Johnson’s house.
That is the extent of the objective facts. None of these facts make it any more likely that Smith was hiding on Johnson’s property rather than in some other location in the surrounding area. Police officers do not obtain probable cause to conduct a search in one place based on the lack of probable cause to search another place. They obtain probable cause because the facts indicate that they will find what they are looking for in the place to be searched. See Terry, 392 U.S. at 22, 88 S.Ct. 1868. No such facts exist in this case.
Moreover, we are at a loss as to why, given the facts before us, Officer Kading reached his decision to search Johnson’s property at all. Officer Kading was chasing Smith in his car; it seems likely that once Smith ran into the woods he would avoid areas where cars could travel. Apparently, this is exactly what happened. After all, this was a sparsely populated rural area. The terrain was hilly and covered with thick brush and trees. Johnson’s property alone consisted of over 12 acres and the neighboring lots were a similar size. Once Smith got into the woods and Officer Kading lost sight of him, Smith’s options were unlimited. He could have run east toward his own property; he could have run southwest past Johnson’s property; he could have stayed in the woods by the highway and used it as a lookout post; or he could have watched Kading drive down Johnson’s driveway and then run north across the street into the woods on the other side of the highway. If he had wanted shelter, Smith could have sought shelter on his parent’s property, his own property, some other neighbor’s property, or out in the woods.
These facts are analogous to those in United States v. Winsor, 816 F.2d 1394 (9th Cir.1987), vacated on other grounds, 846 F.2d 1569 (9th Cir.1988) (en banc). In Winsor, the police followed a bank robber to a “small two-story ‘residential hotel’ ” and watched him “disappear” into the building. Id. at 1395-96. The police went room to room and demanded that the occupants open their doors. Id. at 1396. The police eventually found the person they were looking for. Id.
On appeal, we rejected the government’s argument that the “hot pursuit” exception justified the warrantless search. We stated that inside a hotel, each room enjoys its own zone of protection from unreasonable searches and seizures. Id. at 1397. Based *907on this principle, we held that although the police had probable cause to enter the hotel, they did not have probable cause to search any particular room. Id.
The Appellant in this case argues persuasively that, like the individual hotel rooms in Winsor, all of the homes in the area where the police saw Smith enjoyed a zone of privacy that they could not invade without probable cause. In fact, Johnson presents an even stronger case. In Winsor, 816 F.2d at 1395-96, the police at least knew that the suspect was somewhere in the hotel; they just did not know in which room he was hiding. By contrast, in this case, the area where Smith was last seen is sparsely populated and covered with thick brush and trees. Johnson’s property alone contains over 12 acres and the adjoining property where Smith was last seen is at least as big. The hiding places are potentially endless. This is especially true given that Smith was probably very familiar with the terrain in this area. Furthermore, the police in Winsor caught their man; Smith remains at large.
There is simply no way to transform Officer Kading’s “gut feeling” into “probable cause.” The Supreme Court has clearly stated that gut feelings and inarticulable “hunches” do not equal reasonable suspicion, let alone probable cause. Wardlow, 528 U.S. at 123-24, 120 S.Ct. 673; Terry, 392 U.S. at 27, 88 S.Ct. 1868. Based on this well-established precedent, we hold that probable cause did not exist here.
B.
Our decision would not be altered even if the Government were able to show probable cause. “[N]o amount of probable cause can justify a warrantless search or seizure absent ‘exigent circumstances.’” Horton v. California, 496 U.S. 128, 137 n. 7, 110 S.Ct. 2301, 110 L.Ed.2d 112 (1990). See also LaLonde v. County of Riverside, 204 F.3d 947, 954 (9th Cir.2000). The Government suggests that the exigent circumstances that prevented Deputy Kading from attempting to secure a search warrant was the deputy’s “hot pursuit” of Smith. We hold otherwise.
The hot pursuit exception to the warrant requirement only applies when officers are in “immediate” and “continuous” pursuit of a suspect from the scene of the crime. Welsh, 466 U.S. at 753, 104 S.Ct. 2091; United States v. Salvador, 740 F.2d 752, 758 n. 5 (9th Cir.1984), cert. denied 469 U.S. 1196, 105 S.Ct. 978, 83 L.Ed.2d 980 (1985). In addition, the critical time for determining whether any exigency exists is the moment the officer makes the warrantless entry. They cannot rely on exigencies discovered once they are inside. United States v. Lindsey, 877 F.2d 777, 781 (9th Cir.1989).5
Based on the record in this case, it is clear that the officers’ pursuit of Smith was not “continuous.” After Officer Kad-ing lost sight of Smith- in the woods, he waited for a half hour for backup to arrive. During this time, he returned to Smith’s residence and retrieved a pepper spray canister that he had lost during his confrontation with Smith. When the officers entered Johnson’s property, no one had *908seen Smith for over a half hour. Unless the “continuity” requirement is stretched beyond recognition, the facts of this case simply are not covered by the “hot pursuit” doctrine. The half-hour time period, during which the officers received no new information about where Smith had gone, turned the pursuit from lukewarm to ice cold.
While we respect Deputy Kading’s concern for his safety had he followed Smith into the woods, it does not change the fact that Johnson had not been seen or heard for at least 30 minutes after disappearing into the woods. This is not a case where the police officers always knew exactly where the suspect was, but decided that it would be dangerous for them to enter the property until reinforcements arrive. Cf. Lindsey, 877 F.2d at 779 (noting that when backup arrived, officers saw the suspect through the window of the house). Under such circumstances, the “continuity” of the chase is delayed, but not broken. See United States v. Lindsay, 506 F.2d 166, 173 (D.C.Cir.1974) (stating that “[s]peed and continuous knowledge of the alleged perpetrator’s whereabouts are the elements which underpin this exception to the warrant requirement”).
In this case, however, the continuity of the chase was terminated permanently. Smith did not run into a confined area where Officer Kading could monitor his movements while waiting for his backup to arrive. Smith ran into a wooded area where he was free to run for over a half hour. Once the alleged “pursuit” resumed, the officers no longer had any idea where Smith was.
Under these circumstances, the continuity of the chase was clearly broken and a warrant was required. Although this requirement may be inconvenient to law enforcement, any other outcome renders the concept of “hot pursuit” meaningless and allows the police to conduct warrantless searches while investigating a suspect’s whereabouts.
We also find instructive the Supreme Court’s holding in Welsh v. Wisconsin, 466 U.S. 740, 750, 104 S.Ct. 2091, 80 L.Ed.2d 732 (1984), that “an important factor to be considered when determining whether any exigency, exists is the gravity of the underlying offense for which the arrest is being made.” Id. at 753, 104 S.Ct. 2091. Although the Court did not draw a bright line between felonies and misdemeanors, it cited favorably a number of cases that refused to permit warrantless entries of the home for “nonfelonious crimes.” Id. at 752, 104 S.Ct. 2091. Based on these cases, the Court found that “application of the exigent circumstances exception in the context of a home entry should rarely be sanctioned when there is probable cause to believe that only a minor offense ... has been committed.” Id. at 753, 104 S.Ct. 2091.
Smith was only wanted for misdemeanor offenses. Even his unlawful conduct that gave rise to the initial chase, i.e., resisting arrest, is a misdemeanor under Washington law. See Wash. Rev.Code § 9A.76.040. Although this does not definitely preclude a finding of exigent circumstances, it weighs heavily against it.6
*909Adding another element to this case is the fact that the officers encroached on the property of a person who did not create the exigent circumstances and was completely unrelated to the suspect and his misdemeanors. Here, it was Smith who created the alleged exigency when he resisted arrest. The officers, however, performed a warrantless search of the property of Smith’s neighbor’s home (Johnson). Johnson’s lack of involvement in the situation that created the exigency is another factor weighing against the reasonableness of the warrantless entry.
Very few cases have considered this issue. In National Organization for the Reform of Marijuana Laws v. Mullen, 796 F.2d 276 (9th Cir.1986), we confronted a case where drug enforcement officers, in an attempt to eradicate a local marijuana industry, conducted warrantless searches of neighbor’s properties. We upheld the district court’s determination that “to base probable cause for a search of a neighbor’s home or curtilage on the mere fact that a marijuana garden might be visible from that property is to render meaningless the fourth amendment.” Nat’l Org. for the Reform of Marijuana Laws v. Mullen, 608 F.Supp. 945, 954 (N.D.Cal.1985). Similarly, Johnson’s rights under the Fourth Amendment cannot be waived simply because the officers were hoping to catch a glimpse of Smith on his property.
Of course, the subsequent warrant after the invalid search does not validate the search. Wong Sun v. United States, 371 U.S. 471, 487, 83 S.Ct. 407, 9 L.Ed.2d 441 (1963).
V.
The search was without probable cause and not within the exigent circumstances or hot pursuit exceptions to the Fourth Amendment warrant requirement. Accordingly, we REVERSE on the probable cause issues and REMAND on the curti-lage issue for further proceedings consistent with this opinion.
KOZINSKI, Circuit Judge, with whom Circuit Judges TROTT, T.G. NELSON, SILVERMAN join, Circuit Judge RONALD M. GOULD joins with respects to Parts I, II, III.A and IV, and Circuit Judge PAEZ joins with respect to Part III.A:1
A majority holds that the search was not justified by the hot pursuit exception to the Fourth Amendment, but we need only reach that exception if the search intruded on an area protected by the amendment. Although our prior precedents regarded the curtilage as a factual question, we conclude, by a different majority, that those cases have been overruled by Ornelas v. United States, 517 U.S. 690, 116 S.Ct. 1657, 134 L.Ed.2d 911 (1996). As the historical facts are undisputed, we may consider whether the area outside the shed fell within the curtilage of Johnson’s home under the Fourth Amendment.
I
The chase began when Deputy Chris Fading drove down the driveway leading *910to Steven Dustin Smith’s trailer.2 On spotting the police car, Smith must have guessed that the visit concerned his outstanding arrest warrants because he started to run away. After a brief struggle with the deputy, Smith wriggled out of his jacket and took off down the rural highway, heading west. About halfway between his trailer and Johnson’s homestead, Smith ran south off the road and into a heavily wooded area.
Rather than chasing Smith into the woods, Kading radioed for backup and turned down the driveway to Johnson’s house, where he believed Smith might look for a place to hide. After about 250 yards, Kading reached a locked gate, where he waited until a second officer arrived. The two deputies planned to search for Smith on Johnson’s property before heading into the woods. They pushed in the hasp on the gate and continued more than fifty yards down the driveway to the front of the house. They knocked on Johnson’s front and back doors, but received no answer.
A perimeter fence encircled the buildings on Johnson’s property. Johnson’s home and backyard were surrounded by an interior fence, with the front door opening onto the rest of property. A kennel, surrounded by its own fence, was located about ninety feet from the house, and slightly farther, about 120 to 150 feet from the house, an old shed was built into the side of a hill. The deputies searched a covered area near the back of the house, walked around the outside of the home towards the kennel and looked inside two old vehicles parked nearby. Seeing no sign of Smith, the deputies headed towards the shed. Standing one to two feet in front of its door, Kading smelled the scent of marijuana wafting through a vent. After confirming that the shed was locked, the deputies left the property and continued their search for Smith. They later returned with other deputies and a warrant and discovered 771 marijuana plants in Johnson’s house, shed and kennel.
Before the district court, Johnson moved to exclude his garden from evidence on the ground that the search violated the Fourth Amendment. The government argued that the area around the shed fell outside the curtilage of Johnson’s home and presented evidence in support. Because the district court concluded the deputies were in hot pursuit of Smith, it did not decide whether the area around the shed was within the curtilage. The court noted that “whether I sit here and figure it’s in the curtilage or out of the curtilage,” the deputies’ search was justified because it “in no way intruded on Mr. Johnson’s residence.” Nevertheless, the historical facts underlying the curtilage issue are in the record and undisputed.
II
The Fourth Amendment, as the Supreme Court has interpreted it, permits the police to search all over your land, so long as the officers don’t cross the boundaries of your home. As Justice Holmes observed, “[T]he special protection accorded by the Fourth Amendment to the people in their ‘persons, houses, papers, and effects,’ is not extended to the open fields.” Hester v. United States, 265 U.S. 57, 59, 44 S.Ct. 445, 68 L.Ed. 898 (1924) (quoting U.S. Const. amend. IV). The rationale for this distinction lies not just in the text of the Amendment, but in that provision’s emphasis on preserving privacy. Within *911our homes, we expect to be free from government interference and surveillance. See Oliver v. United States, 466 U.S. 170, 178-79, 104 S.Ct. 1735, 80 L.Ed.2d 214 (1984). But we lack a similar expectation in the open fields of our property. Although the law gives us the right to exclude others from trespassing on our fields, we have no similar privilege to avoid prying eyes while we walk in our front yard. The Supreme Court has read the Fourth Amendment to incorporate these traditional notions by protecting the intimacy of the home, but providing no protection from police searching in the open fields outside the home.
But the 'constitutional boundaries of the home are somewhat larger than the walls of the house; they include the “curtilage,” that “area to which extends the intimate activity associated with the ‘sanctity of a man’s home and the privacies of life.’ ” Id. at 180, 104 S.Ct. 1735 (quoting Boyd v. United States, 116 U.S. 616, 630, 6 S.Ct. 524, 29 L.Ed. 746 (1886)). The concept arises out of the common law, which viewed a man’s home as his “castle of defence.” 4 William Blackstone, Commentaries *225. The curtilage included not only the castle walls, but also those turrets, moats and baileys that adjoined the residence and whose breach the owner would regard as a violation of the security of the home. See id. At the same time, “no distant barn, warehouse, or the like, are under the same privileges.” Id. When we consider modern homes, we might include within the curtilage an adjoining garage, the yard-within the white picket fence, or the gazebo where the kids keep their pool toys. The real question is whether the home owner might reasonably regard those structures as part and parcel of the home itself.
Although there is no mechanical way of determining whether a particular area falls within the curtilage, the Supreme Court, in United States v. Dunn, 480 U.S. 294, 301, 107 S.Ct. 1134, 94 L.Ed.2d 326 (1987), identified four factors that courts should consider in addressing the question: (1) the proximity of the area to the home; (2) whether the area is included within an enclosure surrounding the home; (3) the nature of the uses to which the area is put; and (4) the steps taken by the resident to protect the area from observation by people passing by. These factors do not yield a definite answer; rather they guide courts in determining whether the area is so intimately connected to the home that it should fall under the umbrella of the Fourth Amendment’s protections. See id.
Because the district court did not determine whether the shed fell within the cur-tilage, it failed to conduct the required Dunn analysis. Our ability to review the curtilage question for ourselves depends on whether it is the kind of inquiry susceptible to determination in the first instance by an appellate court, or whether it is an “essentially factual” inquiry requiring us to defer to the district court’s judgment. See United States v. McConney, 728 F.2d 1195, 1203 (9th Cir.1984) (en banc); see also United States v. Depew, 210 F.3d 1061, 1067 (9th Cir.2000) (remanding to the district court for determination of the curti-lage issue). If it is the latter, then we must remand to the district court so that it may make the initial determination.
III
A. In the past, we have joined other circuits in concluding that the Dunn test is factual, to be applied by the district court and reviewed only for clear error. See, e.g., United States v. Traynor, 990 F.2d 1153, 1156-57 (9th Cir.1993); United States v. Reilly, 76 F.3d 1271, 1275 (2d Cir.1996); United States v. Friend, 50 F.3d 548, 552 (8th Cir.1995), vacated on *912other grounds, 517 U.S. 1152, 116 S.Ct. 1538, 134 L.Ed.2d 643 (1996); United States v. Benish, 5 F.3d 20, 24 (3d Cir.1993); United States v. Hatch, 931 F.2d 1478, 1480 (11th Cir.1991). In Traynor, we reasoned that even though the curti-lage question may be viewed as a mixed one of law and fact, the Dunn test was appropriate because the boundaries depended heavily on the factual circumstances of each case. See 990 F.2d at 1156-57; see also United States v. Depew, 8 F.3d 1424, 1426 (9th Cir.1993) (“Every curtilage determination is distinctive and stands or falls on its own unique set of facts.”).
We believe, however, that the Supreme Court’s recent decision in Ornelas, 517 U.S. 690, 116 S.Ct. 1657, 134 L.Ed.2d 911, requires us to reconsider our view of who decides the curtilage issue. In Ornelas, the Court concluded that mixed questions of law and fact under the Fourth Amendment are subject to plenary review by appellate courts. The Court acknowledged that deciding the appropriate standard of review in such instances poses some difficulty. Legal standards such as “probable cause” and “reasonable suspicion” cannot usefully be reduced to clear legal rules, and so they draw their content entirely from their application to concrete cases. See id. at 695-96, 116 S.Ct. 1657. Because such Fourth Amendment questions depend upon a particular factual context, some circuits had concluded (as we had with respect to the curtilage issue) that they must defer to the judgments of the district courts.
However, the Supreme Court decided that such deferential review was inappropriate in Fourth Amendment cases. Although appellate courts must rely on the district courts to determine the historical facts undergirding the Fourth. Amendment inquiry, they need not give the same deference to the district court’s application of the legal standard to those facts. Independent review is necessary for appellate courts to ensure that the Fourth Amendment be applied consistently across cases. See id. at 697, 116 S.Ct. 1657. If appellate courts reviewed such determinations for clear error, then identical factual circumstances could lead to different results depending upon the predilections of each district judge. Likewise, law enforcement officers would lack the “set of rules which, in most instances, makes it possible to reach a correct determination beforehand as to whether an invasion of privacy is justified in the interest of law enforcement.” Ornelas, 517 U.S. at 697-98, 116 S.Ct. 1657 (internal quotation marks and citations omitted).
The considerations catalogued in Ornelas apply with equal force to the determination of where the curtilage ends. There is no conceptual difference between calling an area “curtilage” and telling an officer he had “probable cause” or “reasonable suspicion.” The curtilage question turns on “whether the area in question is so intimately tied to the home itself that it should be placed under the home’s ‘umbrella’ of Fourth Amendment protection.” Dunn, 480 U.S. at 301, 107 S.Ct. 1134. The question of whether an area should be protected by the Fourth Amendment is not ultimately a factual one. It depends upon whether the government’s intrusion in the area “infringes upon the personal and societal values protected by the Fourth Amendment.” Oliver, 466 U.S. at 182-83, 104 S.Ct. 1735. In making that determination, a court must apply this legal value judgment to the facts of each case.3
*913If law enforcement officers are to respect the Fourth Amendment rights of the citizens they serve, they must have the kind of guidance that transcends any one judge’s view of a particular case. The district courts must obviously resolve any controversies over the historical facts underlying the Dunn factors, and appellate courts will review such determinations for clear error. But the application of the law to the facts is not the kind of issue peculiarly within the province of the district courts. Indeed, curtilage questions are easier for appellate courts to review than are determinations of probable cause. The curtilage depends on the layout of the property and the uses to which it is put, factors that are considerably more objective than the inferences drawn from an officer’s experience which underlie the determination of probable cause. See Ornelas, 517 U.S. at 699, 116 S.Ct. 1657 (recognizing that even plenary review of probable cause requires some deference to the expertise of local police). Accordingly, we hold that the determination that a particular search did (or did not) occur within the curtilage must be reviewed de novo on appeal.4
We are the first of the circuits to consider the effect of Ornelas on the curtilage question. The Second Circuit has “assumed, without deciding, that Ornelas requires us to review the district court’s finding of curtilage de novo.” United States v. Reilly, 91 F.3d 331, 331 (2d Cir.1996). Likewise, a prominent Seventh Circuit judge has concluded that “in light of Ornelas, [the curtilage] is a question that we are to decide de novo, without deferring to the district court.” United States v. Redmon, 138 F.3d 1109, 1132 (7th Cir.1998) (en banc) (Posner, C.J., dissenting) (citation omitted).5 The Wisconsin Supreme Court has also read Ornelas as requiring the reviewing court to determine the issue de novo. See State v. Martwick, 231 Wis.2d 801, 604 N.W.2d 552, 557 (2000). No court that has considered Ornelas has ruled to the contrary.
The Supreme Court’s own approach in Dunn supports our plenary review of the curtilage question. There, the Supreme Court held that the barn in question fell outside the curtilage even though the lower courts had not conducted (and of course, could not have conducted) the Dunn analysis. See Dunn, 480 U.S. at 301, 107 S.Ct. 1134. If the Court believed that the curti-lage determination required an on-the-spot judgment by the district court, it could *914have remanded the case for consideration below. Instead, the Court relied on the undisputed facts in the record and determined the question for itself.
B. A majority of the en banc court, after due consideration, holds that the determination of the curtilage must be reviewed de novo. According to Judge Ta-shima, however, this conclusion is merely dicta, because deciding the standard of review is not necessary to the disposition of this case. Judge Tashima advises litigants that the question we purport to answer remains unsettled, and that future panels are free to disagree without violating the law of the circuit.
Judge Tashima’s concurrence raises a fundamental question concerning the development of our circuit law: To what extent is a later panel bound by statements of law contained in opinions of an earlier panel? Judge Tashima would hold that a later panel is free to ignore statements in an earlier opinion-even statements supported by reasoned analysis-if the later panel concludes that the earlier ruling is not necessary to the result reached. Judge Tashima is not the first member of our court to take this position. See, e.g., United States v. Enas, 204 F.3d 915, 920 (9th Cir.2000) (dismissing the legal analysis of an earlier panel as “not necessary to the court’s decision”), reh’g en banc granted, 219 F.3d 1138 (9th Cir.2000). For the reasons explained below, we reject this approach. We hold, instead, that where a panel confronts an issue germane to the eventual resolution of the ease, and resolves it after reasoned consideration in a published opinion, that ruling becomes the law of the circuit, regardless of whether doing so is necessary in some strict logical sense.
The difficulty with Judge Tashima’s approach is that judges often disagree about what is and is not necessary to the resolution of a case. For example, the Enas panel dismissed a key portion of our earlier opinion in Means v. Northern Cheyenne Tribal Court, 154 F.3d 941 (9th Cir.1998), as dicta, because it did not believe the discussion was necessary to the result in Means. See Enas, 204 F.3d at 920. But the judges in Means must have thought they were ruling on an issue that was necessary, else they would not have included it in their opinion. Indeed, a member of the Means panel wrote separately on that very issue-something that would have been wholly superfluous if the Means majority had merely been ruminating.
Panels often confront cases raising multiple issues that could be dispositive, yet they find it appropriate to resolve several, in order to avoid repetition of errors on remand or provide guidance for future cases.6 Or, panels will occasionally find it appropriate to offer alternative rationales for the results they reach.7 Judge Tashi-*915ma did as much in Calderon v. United States District Court (Kelly), 163 F.3d 530, 542 (9th Cir.1998) (en banc), by holding that AEDPA’s statute of limitations didn’t apply because a habeas petition was filed before the statute was enacted and, in the alternative, because equitable tolling applied. Under Judge Tashima’s view, neither rationale would be necessary to the outcome, and future panels could ignore both as dicta.8
If later panels could dismiss the work product of earlier panels quite so easily, much of our circuit law would be put in doubt. No longer would the question be whether an issue was resolved by an earlier panel. Rather, lawyers advising their clients would have to guess whether a later panel will recognize a ruling that is directly on point as also having been necessary. We decline to introduce such uncertainty into the law of our circuit.
We follow, rather, the approach we have taken in deciding whether an issue is “necessarily decided” for purposes of collateral estoppel. As Chief Judge Schroeder explained in United States v. Weems, 49 F.3d 528, 532 (9th Cir.1995), “in order to justify invoking collateral estoppel, a factual determination must have been ‘necessarily’ (and not ‘presumably’) decided in the first proceeding.” But “necessarily,” she noted, means only that the court undeniably decided the issue, not that it was unavoidable for it do so. Over the disagreement of one our colleagues, see id. at 534 (Norris, J., concurring), Weems held that where the court heard evidence and argument from both parties, and specifically ruled on the issue, a party may not escape the ruling’s binding effect on the ground that it was not logically essential to the court’s ultimate determination. See id. at 532.
Of course, not every statement of law in every opinion is binding on later panels. Where it is clear that a statement is made casually and without analysis, where the statement is uttered in passing without due consideration of the alternatives, or where it is merely a prelude to another legal issue that commands the panel’s full attention, it may be appropriate to re-visit the issue in a later case. However, any such reconsideration should be done cautiously and rarely-only where the later panel is convinced that the earlier panel did not make a deliberate decision to adopt the rule of law it announced.9 Where, on *916the other hand, it is clear that a majority of the panel has focused on the legal issue presented by the case before it and made a deliberate decision to resolve the issue, that ruling becomes the law of the circuit and can only be overturned by an en banc court or by the Supreme Court.
A majority of the en banc court has concluded that the determination whether the search took place within the curtilage must be reviewed de novo. The issue was fairly presented to us and refined through the adversary process, and we have decided it after careful analysis, rather than through a casual, off-hand remark or a broad statement of principle. Under these circumstances, our court has undeniably decided the issue, see Weems, 49 F.3d at 532, and our holding becomes the law of the circuit. See United States v. Oshatz, 912 F.2d 534, 540 (2d Cir.1990) (Newman, J.) (“[I]n some contexts expressions of views by an appellate court must be regarded as the law of the circuit, even though not an announcement of a holding or even a necessary step in the reasoning leading to a holding”); United States v. Crawley, 837 F.2d 291, 292-93 (7th Cir.1988) (Posner, J.) (adopting a pragmatic definition of dictum based upon whether the previous panel fully considered the issue and intended for future interpreters to rely on it).10 We disapprove Judge Tashima’s argument to the contrary.
IV
Because five members of our panel would affirm on the record before us, we proceed to examine the Dunn factors based upon the undisputed facts in the record:
1. Proximity. Because the curtilage is defined as the area adjoining and immediately associated with the home, proximity to the home is the first factor under Dunn. In Dunn, the Court noted that 180 feet was a “substantial distance supporting] no inference that the barn should be treated as an adjunct of the house.” 480 U.S. at 302, 107 S.Ct. 1134. Johnson’s shed was located 120 to 150 feet away from the house, not quite as far as in Dunn, but still a substantial distance. We have noted that there is not “any fixed distance at which curtilage ends,” Depew, 8 F.3d at 1427, but our cases have generally regarded the area around a structure 120 feet from the house to lie outside the curtilage. *917Compare United States v. Van Damme, 48 F.3d 461, 464 (9th Cir.1995) (200 feet is outside the curtilage); United States v. Brady, 993 F.2d 177, 178 (9th Cir.1993) (45 feet is outside the curtilage); Traynor, 990 F.2d at 1158 (70 to 75 feet is outside the curtilage); United States v. Calabrese, 825 F.2d 1342, 1350 (9th Cir.1987) (50 feet is outside the curtilage), with United States v. Furrow, 229 F.3d 805, 817 (9th Cir.2000) (100 feet is inside the curtilage); Depew, 8 F.3d at 1427 (60 feet is inside the curti-lage). Thus, standing alone, proximity weighs against finding that the area outside Johnson’s shed lay within the curti-lage.
2. Enclosure. Because, in most cases, the curtilage will be clearly marked, Dunn requires us to examine the fences that demarcated portions of Johnson’s property. Johnson’s shed lies inside the perimeter fence encircling the buildings on his property, but outside an interior fence that includes only the house. Under Dunn, a perimeter fence surrounding a property does not designate the curtilage. See also Traynor, 990 F.2d at 1158. Likewise, Johnson’s perimeter fence, while not surrounding his entire property, encircles an area that contains a dog kennel, the shed, the home and several acres of open space. Given the open fields within the fence, and the existence of several other interior fences, we do not consider the perimeter fence to be relevant in designating the curtilage.
In contrast to the perimeter fence, an interior fence that surrounds the home is a “significant” factor in determining the curtilage. Dunn, 480 U.S. at 302, 107 S.Ct. 1134. Dunn concluded that, where a fence surrounds the home, the curtilage is unlikely to include areas that lie outside that fence. Johnson tries to distinguish this case from Dunn because here, nothing separates the shed from the house; one can walk from the front door of the house to the shed without crossing over any fence. While this is true, it is misleading. Johnson’s house is surrounded on more than 270 degrees by an interior fence that includes a small yard. The only gap in that fence is filled by the house itself. A person walking from the front door to the shed may avoid the fence, but no one could reach the back or sides of the house from the shed without crossing the fence (or entering the house). Thus, the interior fence in this case demarcates an area of land that includes the house but excludes the shed.
3. Use. Dunn also requires us to consider whether the area in question was used or appeared to be used for the intimate activities associated with domestic life and the privacies of the home. Dunn was not entirely clear about whether use is determined by what the officers knew at the time of the search or how the homeowner himself actually used the area. The Court described the factor as “the nature of the uses to which the area is put,” suggesting that the relevant inquiry is the actual use the owner makes of the area. Dunn, 480 U.S. at 301, 107 S.Ct. 1134. At the same time, the Court found it “especially significant” that the officers had objective information that the barn was not being put to domestic use: Aerial photos showed a truck apparently delivering chemicals to the barn, and the officers detected a strong chemical odor emanating from the barn itself. See id. at 302-03, 107 S.Ct. 1134.
Following Dunn, we have emphasized both actual use and objective factors in determining the nature of the uses to which a structure is being put. We have looked at what an observer standing outside the structure could have known. See Brady, 993 F.2d at 178-79 (in determining the use of the building, the police relied *918on the tips of informants, the smell of marijuana and the buzzing of electrical ballasts); Depew, 8 F.3d at 1427 (police received a tip from an informant that property was used to grow marijuana); Traynor, 990 F.2d at 1158 (police were able to smell marijuana and hear the buzzing of electrical ballasts). But we have also considered the actual uses that a homeowner makes of the property. See Van Damme, 48 F.3d at 464 (cultivation of marijuana is not an intimate activity); Brady, 993 F.2d at 178-79 (some evidence that outbuilding was used for storage and as a play area for children, as well as to grow marijuana); Depew, 8 F.3d at 1427 (homeowner was a practicing nudist who liked to walk around in the buff outside his house); Traynor, 990 F.2d at 1158 (no evidence that outbuilding was used for anything other than growing marijuana).
The officers smelled the marijuana while standing in an open area outside the old shed. There is no evidence that Johnson used that area for any intimate household activities, nor that any outside observer would believe the area was so used. Because the officers did not enter the shed, we have no need to decide whether its interior fell within the curtilage. See Van Damme, 48 F.3d at 465 (where the officers smelled marijuana from outside greenhouses, there was no “need to reach the question of whether the interior of the greenhouses constituted ‘open fields’ ”). However, the uses to which the shed was put are relevant to determining whether a person walking from the home to the shed would consider himself as remaining within the curtilage of the home.
Johnson does not claim that he used the old shed for any activities other than growing marijuana. “The cultivation of crops, such as marijuana, is one of those activities that occur in ‘open fields,’ not an intimate activity of the home.” Van Damme, 48 F.3d at 464 (quoting Oliver, 466 U.S. at 179, 104 S.Ct. 1735). Thus, actual use weighs against finding that the area outside the shed fell within the curtilage.
Objective use is somewhat closer. In contrast to Dunn, the deputies had no prior information indicating that Johnson was growing marijuana in the shed or elsewhere on his property. They entered the property looking for Smith. At the same time, a shed is not generally known for housing the intimate activities of domestic life. And Deputy Fading testified that this particular shed was old, lay down a hill from the home and smelled of the marijuana growing inside it. These factors would allow the officers to conclude that the old shed was not considered to be part of the home. Thus, the objective evidence also tilts against finding that the area outside the shed was within the curti-lage.
4. Visibility. The fourth Dunn factor looks at the steps the homeowner has taken to prevent those standing in his open fields from viewing activity in the area alleged to be in the curtilage. In Dunn, the Court noted that the chainlink interior fences were used to corral livestock, rather than block visibility.11 See 480 U.S. at 303, 107 S.Ct. 1134. Similarly, nothing in the record suggests Johnson did anything to shield the shed from the view of those standing in his fields. His fences restricted access but they did not block visibility. *919While the locked shed may have sheltered the marijuana garden inside, there was nothing to stop the officers from observing the area outside the shed. Thus, Johnson did not contrive to protect that area’s privacy as one might protect one’s home.
Taken alone, each Dunn factor weighs against finding that the area outside the old shed lay within the curtilage of Johnson’s property. Together, their weight is conclusive. The officers smelled the marijuana while standing in an open area at least 120 feet away from the house and outside the interior fence surrounding the home. The shed lay down a hill from the house, and indeed was hardly visible from Johnson’s residence. There were no domestic activities conducted in the area, and Johnson made no special effort to shield it from view. Our cases have established that where the police stand outside a nonresidential structure lying away from the house and smell the odor of drugs, they are generally not within the curtilage of the home. See Van Damme, 48 F.3d at 465; Brady, 993 F.2d at 179; Traynor, 990 F.2d at 1159; Calabrese, 825 F.2d at 1350; see also Dunn, 480 U.S. at 301, 107 S.Ct. 1134. We have little difficulty in reaching the same conclusion on the facts of this case.
V
Based on the undisputed facts in the record, we conclude that the area outside the shed was not within the curtilage of Johnson’s home. It seems pointless to send this case back to the district judge, who will surely reach the same result. Accordingly, we would affirm.
. These included charges of driving under the influence, driving while his license was suspended, resisting arrest, malicious mischief, and criminal impersonation.
. The relevant testimony from Deputy Fading's direct examination follows:
Q: Where did you think he was going at that point?
A: I thought it seemed logical for him to run directly to his parents' house
What I do is I get in my patrol vehicle, and I drive down to the next road, the next accessible road which I believe may get me closer to him, and drive up the road and encounter a gate.
Q: Now, why did you go further west, which turns out to be Mr. Johnson's driveway? Why do you drive in that direction?
A: Because my gut feeling was he’s not going to double back and go up his mother’s driveway, because he’s thinking I’m thinking he's going to go to his mom’s house. I’m going to go down this direction I think he’s trying to get away from me going this way. I don't think he’s going to make a circle.
Q: But you had just said you thought he was going to his mother's house.
A: Right. Be he doesn't — my gut feeling was to go this direction.
. Johnson refers to this structure as a “mushroom shed” because the previous owners had used the shed to grow mushrooms.
. This does not change our ruling in United States v. Chavez-Vernaza, 844 F.2d 1368, 1374 (9th Cir.1987) that a state exclusionary rule will not trump an application of federal law in a federal court. Our inquiry rather is into the effect state law may have on the subjective expectations of Johnson and the objective expectations of the local sheriffs regarding his property.
. The Supreme Court developed the exigent circumstances exception to the warrant requirement in Warden v. Hayden, 387 U.S. 294, 298, 87 S.Ct. 1642, 18 L.Ed.2d 782 (1967), where it upheld a warrantless search where "the exigencies of the situation made that course imperative.” It expanded on the "hot pursuit" exigency in United States v. Santana, 427 U.S. 38, 42-43, 96 S.Ct. 2406, 49 L.Ed.2d 300 (1976), where it wrote that "hot pursuit means some sort of a chase, but it need not be an extended hue and cry in and about the public streets.” In the context of a fleeing suspect, any fine distinctions between these two doctrines is immaterial.
. We do not suggest that Officer Kading did not have a right to pursue Smith through the public streets after he resisted arrest. The question, rather, is whether the search can continue when it encroaches on a person's Fourth Amendment rights. Put differently, the question is whose interest should yield — a person's right to be free from warrantless intrusions or law enforcement's interest in apprehending a fleeing suspect. Cf. Brinegar, 338 U.S. at 176, 69 S.Ct. 1302. In situations where an officer is truly in hot pursuit and the underlying offense is a felony, the Fourth Amendment usually yields. See United States *909v. Santana, 427 U.S. 38, 42-43, 96 S.Ct. 2406, 49 L.Ed.2d 300 (1976). However, in situations where the underlying offense is only a misdemeanor, law enforcement must yield to the Fourth Amendment in all but the "rarest” cases. Welsh, 466 U.S. at 753, 104 S.Ct. 2091.
. A majority of the en banc court holds that under Ornelas v. United States, 517 U.S. 690, 116 S.Ct. 1657, 134 L.Ed.2d 911 (1996), the determination of the curtilage must be reviewed de novo on appeal, and we overrule our past cases to the contrary. Thus, Part III.A reflects the opinion of the Court.
. The facts are described in greater detail in the original panel opinion. See United States v. Johnson, 207 F.3d 538, 541-43 (9th Cir.2000), vacated and reh'g en banc granted, 227 F.3d 1169 (9th Cir.2000).
. Relying on Ornelas, the Supreme Court recently recognized that de novo review is appropriate in other contexts that call for "the application of a constitutional standard to the *913facts of a particular case," such as whether the review of the district court’s determination of whether a punitive damage award is constitutionally excessive. Cooper Indus. v. Leatherman Tool Group, — U.S. —, 121 S.Ct. 1678, 149 L.Ed.2d 674 (2001) (internal quotation marks omitted). If the Court concludes Ornelas applies outside the Fourth Amendment context, a fortiori it would seem to apply to analogous determinations within the Fourth Amendment framework.
. We overrule the contrary statements in United States v. Furrow, 229 F.3d 805, 816 (9th Cir.2000); United States v. Soliz, 129 F.3d 499, 502 (9th Cir.1997); United States v. Depew, 8 F.3d 1424, 1426 (9th Cir.1993); United States v. Brady, 993 F.2d 177, 178 (9th Cir.1993); and United States v. Traynor, 990 F.2d 1153, 1156-57 (9th Cir.1993). Because the en banc court ultimately remands the case for a determination of whether the search took place within the curtilage, we retain jurisdiction over any subsequent appeals.
. The majority in Redmon did not reach the curtilage question. The Seventh Circuit, without discussion, has cited Ornelas for the proposition that the curtilage is to be reviewed for clear error. See United States v. Shanks, 97 F.3d 977, 979 (1996). We can’t say whether a subsequent panel in the Seventh Circuit will treat the unreasoned citation as having settled the question. To the extent , it does reflect the judgment that Ornelas requires the curtilage to be reviewed for clear error, we respectfully disagree.
. See, e.g., Marbury v. Madison, 5 U.S. (1 Cranch) 137, 2 L.Ed. 60 (1803) (Marbury has a right and a remedy, but the U.S. Supreme Court lacks jurisdiction); United States v. Vallejo, 237 F.3d 1008, 1026 n. 9 (9th Cir.2001) (recognizing four "independent grounds for reversal” and remanding for a new trial); Association of Mexican-American Educators v. California, 231 F.3d 572, 579 (9th Cir.2000) (en banc) (ruling on whether Title VI and VII apply even though the issues were unnecessary to the disposition of the case); United States v. Weems, 49 F.3d 528, 531 (9th Cir.1995) (Schroeder, J.) ("Because we conclude that the government may retry Weems, we also address Weems' contention that the district court erred in admitting evidence....”).
. See, e.g., United States v. Reid, 226 F.3d 1020, 1026 (9th Cir.2000) (Pregerson, J.) (consent to search was invalid because the party lacked the authority to consent and the consent was involuntary); United States v. Adams, 446 F.2d 681, 684 (9th Cir.1971) (court did not err, and in the alternative, error was harmless). This happens in particular where both prongs of the analysis are doubtful. See, e.g., Kleve v. Hill, 243 F.3d *9151149, 2001 WL 257960, at *2 (9th Cir. Mar. 16, 2001) (sustaining conviction on the grounds that state court decision may have changed the law and that, even if it didn’t, defendant was convicted under the reasoning of the state decision).
. Judge Tashima asserts that "[o]f course,” alternative holdings and dicta "are not the same,” but he doesn’t explain why. Tashima Concurrence at 4 n. 5. Under his rationale, which is that everything not necessary to the result is dicta, both alternative holdings are dicta because neither is necessary to the result. We can test this proposition by asking the question: Would the result change if either of the alternative holdings were removed? The answer, of course, is no. Since either could be removed without affecting the result, neither is necessary, and so under Judge Tashima’s reasoning, dicta. Of course, we do not argue that alternative holdings are dicta; to the contrary. But the fact that they are not, only shows the fallacy of Judge Tashi-ma’s approach.
. For instance, in Vera Cruz v. City of Escondido, 139 F.3d 659, 661 (9th Cir.1997), we noted that, where three previous cases had refrained from announcing a legal standard for the use of deadly force, their remarks could not be taken as settling the question. Likewise, in United States v. Crespo de Llano, 830 F.2d 1532, 1541 n. 2 (9th Cir.1987), we noted that an earlier panel had expressly not decided a standard of review, and so its discussion must be regarded as dicta. In both cases, we measured dicta based on whether the earlier panel intended to decide the issue, not whether the discussion was logically nec*916essary to the earlier disposition. Where a panel tells us it's not deciding the question, of course, we take it at its word.
. Judge Tashima's attempt to distinguish these cases is not persuasive. He accepts that Oshatz read the law of the Second Circuit as including a statement unnecessary to the disposition of the case, but he contends that, since the rule involved trial procedure, it would only bind the district courts, and not later circuit panels. But would later panels of the Second Circuit be free to announce a rule of trial procedure contrary to that announced in Oshatz? Of course not; the law of the circuit binds all courts in the circuit, including later court of appeals panels. In order for the Second Circuit to give different guidance to its trial judges, it would have to go en banc.
Judge Tashima quotes only half the sentence in reading Crawley for the proposition that we may ignore dicta because "the passage was unnecessary to the outcome of the earlier case....” Tashima Op. at 921 (quoting Crawley, 837 F.2d at 292). The sentence continues as follows: "and therefore perhaps not as fully considered as it would have been if it were essential to the outcome.” Crawley, 837 F.2d at 292. Crawley defines dicta by the intention of the previous panel, rejecting as inadequate the definitions Judge Tashima relies on. A passage is dicta where it was "not a fully measured judicial pronouncement,” and so "the court was not attempting to change the established standard and ... no reasonable interpreter of our decisions would have thought it was and relied accordingly." Id. at 293. Under Crawley, a measured judicial pronouncement by a majority is not dicta.
. Judge Ferguson's suggestion that a chain-link fence provides the same expectation of privacy as a fence that blocks visibility is directly contrary to the Supreme Court's holding in Dunn, as well as common sense. See Ferguson Op. at 903. Judge Ferguson's concern for economic inequality doesn't change the fact that one can see through a chainlink fence, no matter whether a rich or poor man stands on the other side.