Appellant Douglas Wayne Brown was charged with driving under the influence and other crimes after he was stopped and later arrested at a traffic safety checkpoint, or roadblock, in Cobb County.1 The trial court granted Appellant’s motion to suppress his statements and other evidence resulting from the stop, ruling that the checkpoint violated the Fourth Amendment. The Court of Appeals reversed that ruling in a 4-3 decision. See State v. Brown, 315 Ga. App. 154 (726 SE2d 654) (2012). We granted Appellant’s petition for certiorari, posing the question: “Did the Court of Appeals employ the correct legal analysis in assessing whether the decision to implement the roadblock was made by supervisory personnel rather than field officers, for a legitimate primary purpose?”
As explained below, we reject Appellant’s initial argument that the checkpoint at which he was stopped was unconstitutional because the police sergeant who authorized it was not a “programmatic-level executive.” Appellant draws this argument from Court of Appeals cases that have improperly conflated the “supervisory personnel” requirement for implementing a specific checkpoint, see LaFontaine v. State, 269 Ga. 251, 253 (497 SE2d 367) (1998), and the distinct requirement that a law enforcement agency’s checkpoint program have an appropriate primary purpose other than the general interest in crime control, which requires review at the “programmatic level” and may involve evidence relating to agency policy and practice and policy-makers other than the supervisor who decided to implement the particular checkpoint at issue, see City of Indianapolis v. Edmond, 531 U. S. 32, 48 (121 SCt 447, 148 LE2d 333) (2000). There is no dispute in this case that the Cobb County Police Department’s traffic safety checkpoint policy satisfies the Edmond requirement, and we adhere to LaFontaine’s holding that the decision to implement a particular checkpoint may be made by any authorized supervisor.
However, Appellant’s fallback argument — that LaFontaine’s “supervisory personnel” requirement was not satisfied in this case, rendering the checkpoint at which he was stopped unconstitutional — has merit. As the dissent below correctly recognized, the dispositive issue in this case is whether the police sergeant decided to *788implement the roadblock as a supervisor in advance or as an officer in the field. See Brown, 315 Ga. App. at 161 (Mikell, P. J., dissenting). And as the > dissent correctly concluded, the trial court’s factual determination that the sergeant made the decision while acting as a field officer rather than in advance as a supervisor was supported by some evidence and therefore was not clearly erroneous, and the trial court’s suppression order should have been affirmed on this basis. See id. Accordingly, we reverse the Court of Appeals’ judgment.2
1. (a) In July 2008, the Cobb County Police Department adopted Policy 5.19 on Traffic Safety Checkpoints. The policy’s stated purpose is
to protect the citizens of Cobb County and to monitor and check driver’s licenses, driver condition, vehicle registrations, vehicle equipment, and various other requirements of the Georgia State Motor Vehicle and Traffic Code.
The policy directs that “traffic safety checkpoints will be conducted in compliance with applicable state and federal law... in a manner that will be of minimal inconvenience to the affected motorists,” before reciting and elaborating on the five elements that this Court said in LaFontaine are required for a particular checkpoint to be deemed constitutional.3
Regarding LaFontaine’s “supervisory personnel” requirement, see 269 Ga. at 253, Policy 5.19 says that “[a]ny supervisor has the authority to implement a checkpoint.” The policy specifies, however, that a checkpoint must be “approved by a supervisor prior to implementation” and explains that “[t]he decision to implement the checkpoint [must be] made by supervisory personnel rather than by officers in the field. This includes the time and location of the checkpoint.” Citing Edmond, the policy adds that checkpoints must be implemented “for a ‘legitimate primary purpose’... [and] cannot be for the purpose of a pretext for ‘general crime detection.’ ” Detailed instructions on the proper staffing of checkpoints are also included.4
*789(b) On or about April 6, 2010, precinct commander Captain Charles Cox e-mailed the officers in the precinct and instructed them to conduct directed traffic enforcement on Groover Road in the vicinity of Allgood Road, in response to a citizen complaint about speeding, racing, and littering. Groover Road is an old, narrow, curvy road with no shoulder that runs along the edge of Blackjack Mountain. The next day, April 7, a shift supervisor, Sergeant Andrew Marchetta, sent a corporal to survey Groover Road. The officer reported to Sergeant Marchetta that the road was not conducive to traffic enforcement using speed detection devices, due to limited sight distances.
According to the trial court, Sergeant Marchetta decided on April 9 to set up a traffic safety checkpoint on Groover Road.5 His plan was to stop every vehicle approaching the checkpoint from either direction to check the driver’s license, vehicle registration, and proof of insurance. A nearby sandpit entrance was to be used for follow-up investigations of drivers for whom screening officers developed articulable suspicion or probable cause of a motor vehicle or other offense.
Sergeant Marchetta directed Officer David Smith to assist him with the checkpoint. The checkpoint began on April 9 at approximately 6:45 p.m., while it was still daylight. The checkpoint was marked by two traffic cones, by Sergeant Marchetta’s and Officer Smith’s patrol cars, which were parked with their blue lights on, and by the officers themselves, who were wearing bright yellow vests. The two officers both acted as screeners. Traffic was light, and in the first 20 minutes, only seven vehicles were stopped, resulting in one citation for an invalid tag.
At about 7:05 p.m., Appellant approached the checkpoint in a Nissan Maxima with a pizza delivery sign on it. He initially attempted to drive around the checkpoint, but when Officer Smith yelled at him *790to stop, he did so. Sergeant Marchetta was screening a vehicle in the other lane, and as he finished, Officer Smith told the sergeant that he smelled marijuana in Appellant’s car. Officer Smith then instructed Appellant to pull over to the sandpit entrance.
As Sergeant Marchetta approached Appellant’s parked car, he saw Appellant make a furtive movement with his hands toward the center of the car and ordered Officer Smith to remove Appellant from the car. As Officer Smith did so, Sergeant Marchetta saw a black object in Appellant’s right hand and a large folding knife clipped to his shorts. As Appellant emerged from the car, he reached toward his waist, dropped the object in his hand (which turned out to be a marijuana pipe), and stomped on it. As Officer Smith tried to control him, Appellant began to fight. He continued to struggle even after being handcuffed, at one point reaching for Officer Smith’s service pistol. Officer Smith broke away from Appellant, and Sergeant Marchetta radioed for backup. After considerable effort by both officers, they were able to subdue and arrest Appellant.
(c) In August 2010, Appellant was indicted for driving under the influence of drugs, violation of the Georgia Controlled Substances Act, two counts of obstruction of a law enforcement officer, and attempted removal of a weapon from a peace officer. Appellant filed a motion to suppress on the ground that the checkpoint violated the Fourth Amendment. After an evidentiary hearing, the trial court granted the motion. The trial court acknowledged that as a supervisor, Sergeant Marchetta had the general authority to implement traffic safety checkpoints pursuant to Policy 5.19. However, the court found, “from the evidence presented,” that Sergeant Marchetta decided to implement the checkpoint at which Appellant was stopped “while acting as an officer in the field and that no evidence was presented that it was planned in advance to occur at a specific time.” The court also found that the checkpoint was not adequately staffed as required by law “to safely and efficiently conduct the checkpoint.”
Pursuant to OCGA § 5-7-1 (a) (4), the State appealed the suppression order to the Court of Appeals, which reversed in a 4-3 decision. See Brown, 315 Ga. App. 154. The majority characterized the evidence as uncontroverted and presenting no question regarding witness credibility, and it therefore applied de novo review to the suppression ruling. See id. The majority then rejected the trial court’s finding that “no evidence was presented that [the checkpoint] was planned in advance to occur at a specific time,” asserting that “[t]he record reveals without dispute that [Sergeant] Marchetta decided to implement the roadblock two days before and even sent another officer to survey the road before implementing it.” Id. at 158. The majority added that “officers are not precluded as a matter of law *791from acting simultaneously as a supervisor and a field officer” and held that understaffing alone was insufficient to make a checkpoint unreasonable under the Fourth Amendment. Id. at 159.
Citing Miller v. State, 288 Ga. 286 (702 SE2d 888) (2010), the dissent took issue with the majority’s decision to apply de novo review. See Brown, 315 Ga. App. at 160-161 (Mikell, P. J., dissenting). The dissent argued that the suppression order should be affirmed because the trial court’s finding that Sergeant Marchetta was acting as a field officer rather than a supervisor when he decided to set up the checkpoint was supported by some evidence. See id. at 161-163. However, the dissent agreed with the majority that the amount of staffing at the checkpoint was “irrelevant” to the constitutional analysis. Id. at 163.
We granted Appellant’s petition for certiorari.
2. (a) “The Fourth Amendment imposes limits on search-and-seizure powers in order to prevent arbitrary and oppressive interference by enforcement officials with the privacy and personal security of individuals.” United States v. Martinez-Fuerte, 428 U. S. 543, 554 (96 SCt 3074, 49 LE2d 1116) (1976). The Fourth Amendment to the United States Constitution guarantees “[t]he right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures.” U. S. Const. amend. IV.6 As its text indicates, the “ultimate touchstone of the Fourth Amendment is ‘reasonableness.’ ” Brigham City v. Stuart, 547 U. S. 398, 403 (126 SCt 1943, 164 LE2d 650) (2006). When a driver brings his vehicle to a stop as a result of a request or show of authority by a law enforcement officer, the officer effectively seizes the vehicle and “ ‘everyone in the vehicle,’ the driver and all passengers.” Arizona v. Johnson, 555 U. S. 323, 327 (129 SCt 781, 172 LE2d 694) (2009) (citation omitted). Such a seizure ordinarily is unreasonable, and hence unconstitutional, absent individualized suspicion. See Chandler v. Miller, 520 U. S. 305, 308 (117 SCt 1295, 137 LE2d 513) (1997).7
The U. S. Supreme Court has recognized, however, a narrow exception to the individualized suspicion requirement for vehicle *792stops made pursuant to a “plan embodying explicit, neutral limitations on the conduct of individual officers.” Brown v. Texas, 443 U. S. 47, 51 (99 SCt 2637, 61 LE2d 357) (1979). Under this checkpoint exception, the reasonableness of the initial stop depends not on individualized suspicion that the driver has committed a traffic violation or other wrongdoing, but instead on the balance between the public interest served by the checkpoint program and the right of individuals to personal security free from arbitrary and oppressive interference by government officials. See Brown, 443 U. S. at 50.
Applying this balancing test, the Supreme Court approved checkpoint programs designed to intercept illegal immigrants near the border, see Martinez-Fuerte, 428 U. S. at 558-562; to keep unlicensed drivers and unsafe vehicles off the road, Delaware v. Prouse, 440 U. S. 648, 663 (99 SCt 1391, 59 LE2d 660) (1979); and to remove drunk drivers from behind the wheel, see Mich. Dept. of State Police v. Sitz, 496 U. S. 444, 447 (110 SCt 2481, 110 LE2d 412) (1990). In each case, the Court distinguished the checkpoint program at issue from a regime of suspicionless stops by roving patrols in the pursuit of the same violations. The Court emphasized two basic threats to liberty that could result if all law enforcement officers were given the authority to make suspicionless vehicle stops as they moved around on patrol in the field. See Martinez-Fuerte, 428 U. S. at 554. Accord LaFontaine, 269 Ga. at 253.
First, the Court focused on the risk of arbitrary stops of citizens as they travel, noting the “grave danger that such unreviewable discretion would be abused by some officers in the field.” Martinez-Fuerte, 428 U. S. at 559. As the Court explained in Prouse, “standard-less and unconstrained discretion is the evil the Court has discerned when in previous cases it has insisted that the discretion of the official in the field be circumscribed, at least to some extent.” 440 U. S. at 661. Accord Sitz, 496 U. S. at 454 (quoting Prouse).
Second, the Court recognized the risk of oppressive interference with the rights of law-abiding citizens, examining both the objective intrusion on their privacy and right to free passage and the subjective intrusion on motorists’ sense of personal security. See Martinez-Fuerte, 428 U. S. at 554, 557-558. The checkpoints approved by the Supreme Court involved brief stops, only a few questions, and a visual inspection from the outside of the vehicle. See id. at 558; Sitz, 496 U. S. at 451-452. The Court described the objective intrusion of such seizures as “minimal” in light of their short duration, the low intensity of the investigation, and the Court’s longstanding view that “[o]ne’s expectation of privacy in an automobile and of freedom in its operation are significantly different from the traditional expectation of privacy and freedom in one’s residence.” Martinez-Fuerte, 428 U. S. *793at 561-562. See also Sitz, 496 U. S. at 452; Illinois v. Lidster, 540 U. S. 419, 424 (124 SCt 885, 157 LE2d 843) (2004) (“The Fourth Amendment does not treat a motorist’s car as his castle.”). The Court also found the subjective intrusion on personal security — the potential “fear and surprise engendered in law-abiding motorists by the nature of the stop” — to be considerably less for stops at a clearly identified police checkpoint than for stops by a roving patrol. Id.
“Roving patrols often operate at night on seldom-traveled roads, and their approach may frighten motorists. At traffic checkpoints the motorist can see that other vehicles are being stopped, he can see visible signs of the officers’ authority, and he is much less likely to be frightened or annoyed by the intrusion.”
Martinez-Fuerte, 428 U. S. at 558 (citation omitted).
(b) Attuned to the concerns highlighted in Martinez-Fuerte, Prouse, and Sitz, in 1998 this Court identified five minimum requirements that a particular checkpoint must satisfy to be upheld as constitutional:
A roadblock is satisfactory where [1] the decision to implement the roadblock was made by supervisory personnel rather than the officers in the field; [2] all vehicles are stopped as opposed to random vehicle stops; [3] the delay to motorists is minimal; [4] the roadblock operation is well identified as a police checkpoint; and [5] the “screening” officer’s training and experience is sufficient to qualify him to make an initial determination as to which motorists should be given field tests for intoxication.
LaFontaine, 269 Ga. at 253.
Addressing the danger that “unreviewable discretion would be abused by some officers in the field,” Martinez-Fuerte, 428 U. S. at 559, LaFontaine’s first, second, and fifth requirements ensure that officers cannot implement checkpoints on the fly while out on patrol, stopping vehicles arbitrarily or targeting (potentially for discriminatory or other improper reasons) individual drivers whom the officers wish to stop and question but lack any articulable reason to do so. Instead, a supervisor in advance, rather than an officer while in the field, must make the decision to conduct the checkpoint; all vehicles must be stopped; and competent screeners must be used. Addressing the burden on free travel and the potential for surprise or fear that being stopped by the police might otherwise generate in innocent *794drivers, see Martinez-Fuerte, 428 U. S. at 557-558, LaFontaine’s third and fourth requirements ensure that law-abiding motorists are only briefly delayed and can readily recognize that the stop is due to a police roadblock rather than any specific focus on them.
(c) Two years after we decided LaFontaine, the U. S. Supreme Court added another layer to the constitutional analysis of vehicle checkpoints. In City of Indianapolis v. Edmond, 531 U. S. 32, the Court reviewed the constitutionality of an extensive highway checkpoint program that the city government conceded had the “primary purpose,” or “proximate goal,” of catching drug offenders and interdicting illegal narcotics. Id. at 34 (punctuation omitted). The Court held that to comply with the Fourth Amendment, a checkpoint program must have, in addition to safeguards on the implementation and operation of the checkpoints, a primary purpose other than “the general interest in crime control.” Id. at 40, 48.
The Court described Martinez-Fuerte, Prouse, and Sitz as approving “only limited exceptions to the general rule that a seizure must be accompanied by some measure of individualized suspicion” and noted that “each of the checkpoint programs that we have approved was designed primarily to serve purposes closely related to the problems of policing the border or the necessity of ensuring roadway safety.” Id. at 41. See also id. at 43 (distinguishing the narcotics-interdiction purpose of the Indianapolis checkpoints from “the type of immediate, vehicle-bound threat to life and limb that the sobriety checkpoint in Sitz was designed to eliminate”). The Court acknowledged the magnitude of the social harms created by illegal narcotics trafficking but said:
We decline to suspend the usual requirement of individualized suspicion where the police seek to employ a checkpoint primarily for the ordinary enterprise of investigating crimes. We cannot sanction stops justified only by the generalized and ever-present possibility that interrogation and inspection may reveal that any given motorist has committed some crime.
Id. at 44.
The Court recognized the problems inherent in determining a checkpoint program’s “primary purpose” but concluded that this limitation was critical, because “[w]ithout drawing the line at roadblocks designed primarily to serve the general interest in crime control, the Fourth Amendment would do little to prevent such intrusions from becoming a routine part of American life.” Edmond at *79542. The Court rejected the government’s argument that its narcotics-interdiction checkpoints were justified by their secondary purposes of removing impaired drivers from the road and verifying licenses and registrations, reasoning that without an inquiry into primary purpose, “law enforcement authorities would be able to establish checkpoints for virtually any purpose so long as they also included a license or sobriety check.” Id. at 46. See also id. at 42 (emphasizing the need for a “check on the ability of the authorities to construct roadblocks for almost any conceivable law enforcement purpose”).8
“For this reason,” the Court said, “we examine the available evidence to determine the primary purpose of the checkpoint program.” Edmond at 46. The Court reiterated, however, that “ ‘[subjective intentions play no role in ordinary, probable-cause Fourth Amendment analysis,’ ” id. at 45 (citation omitted), and cautioned that “the purpose inquiry in this context is to be conducted only at the programmatic level and is not an invitation to probe the minds of individual officers acting at the scene,” id. at 48. Examining the Indianapolis narcotics checkpoint program at this “programmatic level,” the Court found that its “primary purpose ... is ultimately indistinguishable from the general interest in crime control” and therefore held that “the checkpoints violate the Fourth Amendment.” Id.9
*796(d) Thus, Edmond supplemented LaFontainés requirements for the implementation and operation of a constitutionally valid checkpoint by adding the requirement of an inquiry into the primary purpose of the checkpoint program10 LaFontaine’s requirement that the decision to implement a particular roadblock be made by “supervisory personnel” is distinct from Edmond’s requirement that the roadblock program have a primary purpose other than the general interest in crime control. They involve different factual inquiries, and they serve different objectives in the Fourth Amendment scheme.
The first LaFontaine requirement ensures that a police checkpoint cannot be set up at the “standardless and unconstrained discretion ... of the official in the field,” Prouse, 440 U. S. at 661; the decision as to where and when to conduct a checkpoint must instead be made in advance by a supervisor, removing such discretion from every officer patrolling out in the field.11 By contrast, Edmond requires an examination of the policy purpose of the checkpoints, viewed “at the programmatic level,” to ensure that an agency’s checkpoints are established primarily for a lawful and focused purpose like traffic safety rather than to detect evidence of ordinary criminal wrongdoing. Edmond, 531 U. S. at 48. See also id. at 47-48 (distinguishing between the “general scheme” of an agency’s checkpoints and the intentions “of individual officers acting at the scene” of *797a particular checkpoint).
Put another way, the focus of the Edmond requirement is not on when, where, how, and by whom the specific checkpoint was implemented and operated — the focus of the LaFontaine requirements — but rather on why the agency uses checkpoints. Thus, if the primary purpose of the checkpoint program is crime-fighting in general, then the checkpoints implemented under that program are unconstitutional, even if the decision to implement them was made well in advance by the official with the most policymaking authority in the agency — as was true of the Indianapolis narcotics checkpoint program at issue in Edmond. See 531 U. S. at 35 (noting that the checkpoints were operated pursuant to written directives from the chief of police and that checkpoint locations were selected weeks in advance). Conversely, if the agency has established a checkpoint policy with an approved primary purpose other than general crime control, how that program is implemented at a particular time and place is a question of the LaFontaine requirements, not Edmond.
In sum, the five LaFontaine requirements and Edmond’s appropriate primary purpose requirement address different issues in the Fourth Amendment analysis and should be addressed separately by courts reviewing disputed checkpoint stops — as this Court has done after Edmond. See Hardin v. State, 277 Ga. 242, 243-244 (587 SE2d 634) (2003) (describing Edmond as prohibiting the use of roadblocks “primarily to uncover evidence of ‘ordinary criminal wrongdoing’ ” and noting that the “[e]vidence at trial showed that the roadblocks in this case . . . were implemented as safety checkpoints based on a number of citizen complaints about drunken and reckless drivers in the area” before examining whether “the roadblock at which [Hardin] was stopped violated the LaFontaine standards”). Cf. Phillips v. State, 287 Ga. 560, 562-563 (697 SE2d 818) (2010) (examining only the challenged LaFontaine requirements, where the defendant apparently did not claim that the checkpoint program lacked a primary purpose other than general crime control); Lutz, 274 Ga. at 74 (same).
It should also be noted that compliance with the Edmond and LaFontaine requirements does not necessarily end the Fourth Amendment analysis of a checkpoint case. The ultimate question remains whether, under the totality of the circumstances, the challenged stop was reasonable. See Baker, 252 Ga. App. at 701. Thus, even if the checkpoint program and the particular checkpoint at which the driver was stopped satisfy the formal requirements, the reviewing court may consider evidence that the checkpoint basis for the stop was pretextual, see LaFontaine, 269 Ga. at 253, or that the checkpoint was used to harass, see Martinez-Fuerte, 428 U. S. at 567 n. 19, *798or was otherwise arbitrary or oppressive. See Lidster, 540 U. S. at 426 (“And, of course, the Fourth Amendment’s normal insistence that the stop be reasonable in context will still provide an important legal limitation on police use of this kind of information-seeking checkpoint.”).
(e) About a year after Edmond was decided, our Court of Appeals considered, in a whole-court case, how the U. S. Supreme Court’s new police checkpoint decision affected the test this Court had established in LaFontaine. See Baker, 252 Ga. App. at 697-703 (whole court). The Baker majority recognized that Edmond had altered the constitutional analysis, but it erred in failing to recognize that Edmond supplemented LaFontaine’s five requirements for a specific checkpoint to be deemed lawful with a new and distinct focus on the agency’s checkpoint program. Instead, the majority held that the first LaFontaine requirement had been “modified by Edmond,” Baker, 252 Ga. App. at 702 (emphasis added), so that the State now needed to prove both “that the decision to implement the checkpoint in question was made by supervisory officers and not officers in the field and that the supervisors had a legitimate primary purpose.” Id. (emphasis in original).12
Baker’s merger of the Edmond requirement into the first LaFontaine requirement has created confusion in some cases, because the official who decided to implement the specific checkpoint at issue — the focus of the LaFontaine inquiry — may not be, and need not be, the official or officials whose primary purpose for establishing the agency’s checkpoint program must be determined under Edmond. As discussed previously, in accord with the U. S. Supreme Court’s pre-Edmond checkpoint cases, LaFontaine’s first requirement works to control the potential for arbitrary stops that would exist if every officer in the field could implement a checkpoint by moving the implementation decision up and away from the field to a supervisor. If that supervisor is low-ranking, he may have the authority to implement a particular vehicle checkpoint under LaFontaine and his agency’s checkpoint program, and the primary purpose of that particular checkpoint may be appropriate.13 But a low-level supervisor may not always be aware of or able to testify to the primary purpose of the agency’s overall checkpoint program, which may have been *799established and may be overseen by officials at higher levels of the agency. Under Edmond, it is at that “programmatic level,” 531 U. S. at 48, that the “primary purpose” inquiry must focus, with the goal of ensuring that the agency has not authorized vehicle checkpoints primarily for general crime control but rather for an appropriately limited purpose like traffic safety.14
When not conflated, the first LaFontaine requirement and Edmond’s appropriate primary purpose requirement can be more easily understood and applied. Two distinct questions are presented:
(1) Was the police checkpoint at issue implemented pursuant to a checkpoint program that had, when viewed at that programmatic level, an appropriate primary purpose other than general crime control?
(2) If so, was the decision to implement that specific checkpoint made by a supervisor in advance rather than by an officer in the field?
To the extent that Baker and subsequent Court of Appeals’ cases conflate these issues, those cases are disapproved.
(f) To.summarize, where a defendant challenges his initial stop at a police checkpoint by way of a motion to suppress, the State bears the bur den of proving that the seizure was constitutional. SeeOCGA § 17-5-30 (b); Baker, 252 Ga. App. at 699. This requires the State to prove that the stop was reasonable under the totality of the circumstances. See id. at 701. At a minimum, the State must show that the law enforcement agency’s checkpoint program had an appropriate primary purpose other than ordinary crime control — a purpose examined at that programmatic level, rather than by trying to determine the motives of the supervisor who implemented and the officers who conducted the particular checkpoint at issue. See Edmond, 531 U. S. at 48. The State must also prove that the particular *800checkpoint at which the defendant was stopped was properly implemented and operated — that the five requirements enumerated in LaFontaine were met. See 269 Ga. at 253.
3. We now apply these legal principles to the facts of Appellant’s case.
(a) This case involves no Edmond issue. Appellant does not dispute that the Cobb County Police Department’s checkpoint policy, viewed at the programmatic level, has a primary purpose other than the general interest in crime control, namely, “to monitor and check driver’s licenses, driver condition, vehicle registrations, vehicle equipment, and various other requirements of the Georgia State Motor Vehicle and Traffic Code.” See Prouse, 440 U. S. at 663 (approving checkpoints for the purposes of removing unlicensed drivers and unsafe vehicles from the road); Sitz, 496 U. S. at 447 (approving checkpoints for the purpose of removing impaired drivers from the road). Indeed, Policy 5.19 expressly forbids the use of roadblocks as a pretext for general crime detection. Appellant also does not challenge the State’s evidence that the Groover Road checkpoint at which he was stopped complied with the final four LaFontaine requirements, that is, that all vehicles were stopped, the delay to motorists was minimal, the operation was well identified as a police checkpoint, and the screening officers’ training and experience qualified them to decide whether motorists should be given field tests for intoxication. See LaFontaine, 269 Ga. at 253. Finally, Appellant does not argue that his stop was unreasonable under the totality of the circumstances.
(b) Thus, if Appellant is to prevail, he must do so based on his contention that the State failed to prove that the first LaFontaine requirement was satisfied — that “the decision to implement the roadblock was made by supervisory personnel rather than the officers in the field.” Id. Appellant advances two arguments on this point. First, he maintains that Sergeant Marchetta did not qualify as “supervisory personnel” under LaFontaine because the State failed to prove that the sergeant was a “programmatic-level executive.” Second, Appellant claims that the State failed to prove that Sergeant Marchetta made the decision to implement the checkpoint in advance while acting in a supervisory role, instead of while acting as an officer in the field. The first argument is meritless, but the second is correct.
(1) Appellant does not deny that Sergeant Marchetta was a shift supervisor in the Cobb County Police Department and that Policy 5.19 gave him the authority to implement a traffic safety checkpoint. Instead, Appellant asserts that to satisfy LaFontaine’s requirement that the decision to implement a checkpoint be made by “supervisory *801personnel,” the State also was required to prove that Sergeant Marchetta was a “programmatic-level” executive or supervisor. However, “supervisory personnel” simply means “supervisors.” See, e.g., LaFontaine, 269 Ga. at 252-253 (holding that the “supervisory personnel” requirement was satisfied because the decision to implement the roadblock was made by a “State Patrol supervisor”); Brent, 270 Ga. at 161-162 (using the terms “supervisory personnel” and “supervisor” interchangeably); Baker, 252 Ga. App. at 698 (same). As explained in Division 2 (b) above, the purpose of the first LaFontaine requirement is not to require approval of a particular roadblock at some high level of agency management, but simply to ensure that the implementation decision is made above and away from “field officers” on roving patrol, whose unfettered exercise of discretion is feared. Requiring such decisions to be made two (or more) levels away from the field, rather than one, would not make a significant difference in this respect.
Moreover, as explained above in Division 2 (d), the term “programmatic level” relates to the Edmond requirement that a checkpoint program have a primary purpose other than ordinary crime control, a requirement distinct from the LaFontaine requirements. Appellant has cited no decision from the U. S. Supreme Court or this Court that holds that an “executive” or “programmatic-level” law enforcement official must make the decision to implement a particular checkpoint in order for that checkpoint to be constitutional, as opposed simply to a supervisory official who has the authority to implement checkpoints and does so in his capacity as a supervisor rather than as an officer out in the field. Edmond did not use the term “executive” or “programmatic level” in this context. The only reference to “ ‘executive-level officers’ ” in LaFontaine was in the partial dissent, see 269 Ga. at 257 (Sears, J., concurring in part and dissenting in part) (citing what is now 5 LaFave § 10.8 (d)), and the author of that partial dissent soon thereafter agreed that supervisory approval of a checkpoint is sufficient, see Brent, 270 Ga. at 163 (Sears, J., concurring). See also 5 LaFave § 10.8 (d) (using the terms “executive-level officers” and “supervisory officials” interchangeably). And while the dissent below, after a discussion that conflated the Edmond and LaFontaine requirements as in previous Court of Appeals cases, employed the phrase “ ‘executive level programmatic’ official,” even it immediately made clear that the phrase was being used as to the first LaFontaine requirement simply to describe any “non-field officer.” Brown, 315 Ga. App. at 161-163 (Mikell, P. J., dissenting). In short, *802precedent does not support the imposition of the requirement that Appellant advocates.
Indeed, it is unclear how courts would apply such a requirement. If “executive” or “programmatic-level” means something other than “supervisor” or “non-field officer,” Appellant does not explain how to accurately identify such officials, suggesting only that it would require a complicated, multi-factor analysis of the official’s rank, duties, management responsibilities, and decision-making authority in terms of hiring, firing, pay rates, and work scheduling. Nor does he explain how such a requirement would be applied consistently to the multitude of law enforcement agencies throughout Georgia. Who qualifies as an “executive”-level official in a large metropolitan police department with hundreds of sworn officers and multiple layers of management might be quite different from who qualifies as an “executive”-level official in a small municipal agency with only a handful of officers, and even agencies of the same size may be organized in many different ways. Courts are not in the business of micromanaging how law enforcement agencies should organize themselves and delegate authority.
To be sure, thorough oversight of the implementation of specific checkpoints by higher-level officials in law enforcement agencies is a commendable practice and one that may well identify and deter problems in checkpoint programs — including, potentially, implementation issues that could lead a court reviewing the checkpoint program to question whether it really has an appropriate primary purpose. But we see no basis to depart from precedent and impose such an ill-defined requirement as a purported constitutional obligation. Accordingly, we reject Appellant’s argument that an officer cannot be classified as “supervisory personnel” under the first LaFontaine requirement unless he exercises “programmatic-level” authority within his agency or is identified as an “executive.”
(2) After asking us to revise the first LaFontaine requirement, Appellant asks us simply to enforce it, arguing that the State failed to prove that Sergeant Marchetta made the decision to implement the checkpoint on Groover Road as a supervisor in advance rather than as an officer in the field. Supported by the dissent below, Appellant contends that the trial court’s factual finding that Sergeant Marchetta decided to implement the checkpoint “while acting as an officer in the field” was supported by some evidence at the suppression hearing and therefore should have been affirmed. We agree.
In Miller v. State, 288 Ga. 286 (702 SE2d 888) (2010), this Court reiterated “three fundamental principles which must be followed *803when conducting an appellate review” of a trial court’s ruling on a motion to suppress. Id. at 286.
First, when a motion to suppress is heard by the trial judge, that judge sits as the trier of facts. The trial judge hears the evidence, and his findings based upon conflicting evidence are analogous to the verdict of a jury and should not be disturbed by a reviewing court if there is any evidence to support [them]. Second, the trial court’s decision with regard to questions of fact and credibility must be accepted unless clearly erroneous. Third, the reviewing court must construe the evidence most favorably to the upholding of the trial court’s findings and judgment.
Id. (citation omitted). These principles apply equally whether the trial court ruled in favor of the State or the defendant. See id. at 286-287. The majority opinion below deviated from these principles in reversing the trial court’s suppression order.
The Court of Appeals majority asserted that de novo review of the trial court’s order was appropriate because “[t]he record reveals without dispute that Marchetta decided to implement the roadblock two days before” it occurred. Brown, 315 Ga. App. at 158. That is incorrect. It is true that on April 7, 2010, two days before the checkpoint was set up, a corporal reported to Sergeant Marchetta that Groover Road was not conducive to speed detection devices, so his attention may have turned at that time to using a roadblock at some point for the directed traffic enforcement that Captain Cox had requested.
But when the decision was made to actually implement a roadblock on Groover Road at 6:45p.m. on April 9 was not undisputed. At the suppression hearing, Sergeant Marchetta testified that he informed Captain Cox on either April 8 or 9 about his plan to set up a checkpoint on Groover Road, although he did not claim to have specified when the checkpoint would be set up, and when Captain Cox testified, he squarely denied that he had any advance notice from Sergeant Marchetta of the checkpoint at which Appellant was stopped. In addition, the other officer who participated in the checkpoint, Officer Smith, testified at the hearing, and one reasonable interpretation of his testimony is that Sergeant Marchetta did not inform him of the planned checkpoint until Officer Smith arrived at the scene. Furthermore, Sergeant Marchetta testified that he did not order the checkpoint in writing, and the State presented no documents at the hearing showing that Sergeant Marchetta (or any other supervisor) decided on the time that the checkpoint would be established prior to *804its implementation in the field. Law enforcement agencies that fail to require and maintain documentation as to when, where, and by whom the decision to implement a checkpoint is made may find it harder to prove those facts if disputed in court many months later. See Brown, 315 Ga. App. at 162 n. 17 (Mikell, P. J., dissenting) (“Evidence that the key questions of time, place, and duration were settled back at headquarters in the supervisor’s office would be quite persuasive.”).
Thus, the relevant evidence was not undisputed, and the principles set forth in Miller apply. Under those principles, the record supports the trial court’s express factual findings that Sergeant Marchetta made the decision to implement the checkpoint while acting as an officer in the field and that the State failed to prove that the checkpoint was planned in advance to occur at a specific time. Indeed, even putting aside the testimony that conflicted with Sergeant Marchetta’s account, the trial court had the opportunity to observe his testimony firsthand and was entitled to disbelieve his claim that he decided to implement the checkpoint in advance. As we explained in the context of a motion to suppress in Tate v. State, 264 Ga. 53 (440 SE2d 646) (1994),
[t]he trier of fact is not obligated to believe a witness even if the testimony is uncontradicted and may accept or reject any portion of the testimony. . . . [Thus,] a rational trier of fact can choose to reject even “undisputed” testimony .... Factors such as demeanor, contradictory or inconsistent statements and evidence that an officer had “ulterior motives” can all lead a finder of fact to disregard testimony by an officer....
Id. at 56 & n. 5.
Based on its adequately supported finding that the State had not proved compliance with the first LaFontaine requirement, the trial court correctly concluded that the checkpoint at which Appellant was stopped violated the Fourth Amendment. Consequently, the Court of Appeals erred in reversing the trial court’s suppression ruling.
(c) The trial court cited as an additional ground for its suppression ruling the fact that the Groover Road checkpoint was not sufficiently staffed “to safely and efficiently conduct the checkpoint,” noting Sergeant Marchetta’s testimony that he simultaneously acted as the checkpoint supervisor, an initial screener, a follow-up investigator of drivers whom the screeners developed reasonable suspicion or probable cause to detain, and the back-up officer to chase motorists who refused to stop at the checkpoint. The trial court indicated that *805the roadblock violated the standards and procedures of Policy 5.19 and concluded that “[t]he roadblock was not sufficiently manned, as required by LaFontaine."
The Court of Appeals unanimously rejected this alternative ground for suppression, see Brown, 315 Ga. App. at 159; id. at 163 (Mikell, P. J., dissenting), and we agree that the trial court erred in ruling that the understaffing of the checkpoint was, in and of itself, a violation of LaFontaine and the Fourth Amendment. Violations of police policies like Policy 5.19 do not, ipso facto, violate the Constitution; law enforcement agencies can (and often commendably do) establish policies intended to avoid coming close to the constitutional line. And LaFontaine does not require that a specific number of officers participate in a checkpoint.
Contrary to the Court of Appeals dissent’s view, however, insufficient staffing of a checkpoint is not “irrelevant” to the constitutional analysis, Brown, 315 Ga. App. at 163 (Mikell, P. J., dissenting), and indeed it may be relevant in evaluating all of the LaFontaine requirements. If staffing is so limited, as it was here, that the officer who decides to implement the checkpoint must also play an integral role in its operation, that fact may raise questions about when and in what capacity the officer made the implementation decision. See Thomas v. State, 277 Ga. App. 88, 90, 92 (625 SE2d 455) (2005) (concluding that the first LaFontaine requirement was not met where the decision to conduct the roadblock was made by a senior patrol officer “supervising the other officers on his shift . . . while he spoke with [them] at a convenience store immediately before implementing the roadblock itself” and the senior officer “actively participated in all aspects of the roadblock as did the other field officers”).15
*806Understaffing may also make it impossible to stop all vehicles while keeping the delay to law-abiding motorists minimal. The presence of only a few police officers and vehicles may make the operation less readily identifiable as a police checkpoint. And without a sufficient number of officers, those with the training and experience in detecting impaired drivers necessary to act as screeners may not be available at all times to screen drivers, due to the necessity of assisting other officers if a potentially dangerous situation develops, as it did in this case. Finally, adequate staffing ensures that the agency is devoting resources to the asserted primary purpose of the checkpoint, while also putting a natural limit on the number of checkpoints an agency will utilize. See Lidster, 540 U. S. at 426 (observing that in addition to Fourth Amendment constraints, “practical considerations — namely, limited police resources and community hostility to related traffic tie-ups — seem likely to inhibit” any unreasonable proliferation of police checkpoints).
Thus, the two-officer staffing of the checkpoint in this case was not in itself a constitutional violation, but it certainly did not enhance the State’s ability to show that the checkpoint was implemented and operated lawfully. In addition, the questionable application or violation of the staffing requirements set out in the police department’s checkpoint policy could be considered by the trial court in evaluating the credibility of Sergeant Marchetta’s testimony that he decided to implement the checkpoint two days in advance.
4. Traffic safety checkpoints can be a valid and important means of law enforcement. However, police checkpoint programs must have an appropriate primary purpose other than general crime control, and each checkpoint must also be implemented and operated so as to control the risks of unconstrained discretion that “would be abused by some officers in the field,” Martinez-Fuerte, 428 U. S. at 559, and of “oppressive interference by enforcement officials with the privacy and personal security of individuals,” id. at 554. While we reject Appellant’s request that we amend the well-established safeguards that the U. S. Supreme Court and this Court have required to ensure that checkpoints comply with the Fourth Amendment, we reiterate those requirements, under which Appellant’s stop at the checkpoint on Groover Road was unconstitutional. Our holding follows from the factual findings of the trial court, which are supported by the record. Indeed, we rely on the trial courts of Georgia to evaluate the evidence presented in each case to determine if the requirements for a constitutionally valid checkpoint have been satisfied.
Judgment reversed.
Thompson, C. J., Hines, P. J., Benham, Hunstein, Melton, JJ., and Judge Jerry W. Baxter concur. Blackwell, J., disqualified. *807Decided October 21, 2013. John A. Steakley, for appellant. D. Victor Reynolds, District Attorney, John R. Edwards, Richele P. Anderson, Daniel J. Quinn, Assistant District Attorneys, for appellee.The terms “checkpoint” and “roadblock” are generally used interchangeably to refer to situations where law enforcement officers require vehicles passing on a road to stop at a designated point to allow brief checks of the vehicles and their occupants. This opinion will do the same.
We have considered the legal issues in this case alongside those in another police checkpoint case, which we also decide today, in which we focus further on the Edmond requirement. See Williams v. State, 293 Ga. 883 (750 SE2d 355) (2013).
The LaFontaine requirements are listed and discussed in Division 2 (b) below.
Policy 5.19 addresses staffing as follows:
A. Supervisor Approval
. . . [T]he approving supervisor should ensure:
1. There is adequate manpower to safely and efficiently conduct the checkpoint. Normally, the checkpoint supervisor should not act as a screening officer....
*789B. Personnel
The traffic volume will in most cases determine the number of personnel needed for the traffic safety checkpoint. At a minimum:
1. Officers will be needed to screen motorists in the roadway. Only officers with the appropriate training and experience should be deployed as screeners.
2. Officers will be needed to conduct follow-up investigations of motorists whom the screeners have developed articulable suspicion or probable cause to detain.
3. At least one additional officer will be needed to initiate traffic stops of motorists who violate state law in the vicinity of the traffic safety checkpoint and/or refuse to stop at the checkpoint.
As discussed further in Division 3 (b) (2) below, the trial court found that Sergeant Marchetta made the decision to implement the checkpoint on the day it was set up (April 9), while the Court of Appeals majority said that the decision was made two days in advance (on April 7).
The Georgia Constitution’s Bill of Rights contains the same language. See Ga. Const. of 1983, Art. I, Sec. I, Par. XIII. Appellant raises no separate claim under the Georgia Constitution, but in any event we have held that Paragraph XIII is applied in accord with the Fourth Amendment in this context. See Brent v. State, 270 Ga. 160, 162 (510 SE2d 14) (1998).
The quantum of individualized suspicion required depends on the context. Typically the standard is probable cause to believe that the individual seized has committed a crime, but more limited seizures may be based on a lesser showing of articulable suspicion, based on specific, objective facts, that a person has been, is, or is about to be engaged in criminal activity. See Bailey v. United States, 568 U. S._,_(133 SCt 1031, 1037, 185 LE2d 19) (2013); United States v. Hensley, 469 U. S. 221, 227 (105 SCt 675, 83 LE2d 604) (1985).
The Court distinguished situations where “the primary purpose would otherwise, but for some emergency, relate to ordinary crime control,” such as “an appropriately tailored roadblock set up to thwart an imminent terrorist attack or to catch a dangerous criminal who is likely to flee by way of a particular route.” Edmond at 44. The Court also reserved for another day “whether the State may establish a checkpoint program with the primary purpose of checking licenses or driver sobriety and a secondary purpose of interdicting narcotics,” and “whether police may expand the scope of a license or sobriety checkpoint seizure in order to detect the presence of drugs in a stopped car.” Id. at 47 n. 2. See also Lidster, 540 U. S. at 423-426 (holding that Edmond did not govern “information-seeking” vehicle stops at a highway checkpoint set up to ask drivers for information as potential witnesses to a recent hit-and-run accident on the same road). This case does not require us to address these issues.
Edmond’s requirement that a checkpoint program have a primary purpose beyond detecting “evidence of ordinary criminal wrongdoing” or advancing “the general interest in crime control,” 531 U. S. at 42, has sometimes been shorthanded as requiring simply a “legitimate primary purpose.” See, e.g., Baker v. State, 252 Ga. App. 695, 702 (556 SE2d 892) (2001) (whole court). That phrase is misleading. After all, the primary purpose of the checkpoint program in Edmond was interdiction of illegal narcotics, which obviously is a ‘legitimate” law enforcement goal. Edmond’s focus is not on whether the checkpoint program has a proper law enforcement objective, but rather on whether that objective is “ ‘distinguishable from the general interest in crime control.’ ” Edmond, 531 U. S. at 40 (quoting Prouse). See also id. at 42 (explaining that “questions concerning what means law enforcement officers may employ to pursue a given purpose” such as illegal narcotics interdiction cannot be based solely on the gravity of the threat (emphasis added)). Accordingly, we will refer to the “Edmond requirement” or the “appropriate primary purpose requirement” - with “appropriate” understood to mean a primary purpose that is both legitimate for law enforcement to pursue and can be distinguished from general crime control.
In LaFontaine, this Court examined the primary purpose of the individual checkpoint at issue, finding that it was implemented for the proper purpose of “routine traffic checks.” 269 Ga. at 253. However, the State Patrol’s checkpoint program may have authorized roadblocks “in order to detect ‘any violation of Georgia law, whatsoever,’ ” id. at 254 (Sears, J., concurring in part and dissenting in part), in which case its primary purpose, viewed at the programmatic level, could be seen as general crime control. Accordingly, the checkpoint approved in LaFontaine might be deemed invalid after Edmond. See Baker, 252 Ga. App. at 698 n. 4 (noting this point).
In LaFontaine, this Court upheld the constitutionality of a roadblock on Old Atlanta Road in Forsyth County where “[t]he decision to implement the roadblock was made by a State Patrol supervisor” but “the determination as to the location of the roadblock was made by the field officers.” 269 Ga. at 252 (emphasis added). It is not clear from the opinion whether the field officers’ discretion regarding “location” was limited to choosing the particular spot on Old Atlanta Road to set up the roadblock or instead extended to establishing a roadblock anywhere in the field officers’ territory. Later cases, however, have construed LaFontaine’s reference to location narrowly, which we agree is necessary to ensure that field officers are not vested with overbroad discretion regarding this important aspect of roadblocks. See, e.g., Lutz v. State, 274 Ga. 71, 74 (548 SE2d 323) (2001) (noting that “[t]he decision of where and when to place the roadblock was made by supervisory personnel rather than field officers”); Coursey v. State, 295 Ga. App. 476, 477 (672 SE2d 456) (2009) (reading LaFontaine to require that “supervisory officers decide[ ] where and when to implement” a particular checkpoint). See also 5 Wayne R. LaFave, Search and Seizure § 10.8 (a) (5th ed. 2013) (hereinafter LaFave) (“[I]t is possible that a roadblock purportedly established to check licenses would be located and conducted in such a way as to facilitate the detection of crimes unrelated to licensing. That risk can be minimized by a requirement that the location of such a roadblock be determined by a supervisory official. . ..” (footnote omitted)).
Beyond this passage’s analytical defect, it also appears to be the place where Georgia courts began summarizing Edmond as requiring a police checkpoint to have simply a “legitimate primary purpose.” As discussed in footnote 9 above, that phrasing is misleading.
Of course, if the primary purpose of the checkpoint at issue were inconsistent with the agency’s checkpoint program, the checkpoint could not be justified by the program and would be invalid (assuming it was not justified by an emergency situation, see footnote 8 above).
Judge Eldridge’s dissent in Baker recognized this distinction between LaFontaine and Edmond-.
In fact, Edmond does not “modify” LaFontaine, as the majority asserts. The two cases have little to do with each other, except that both deal with roadblocks. Nonarbitrary interference á la the LaFontaine procedural criteria may be shown, yet a permissible purpose for a roadblock still goes unproved. Likewise, a constitutionally permissible purpose for a roadblock maybe shown, while a nonarbitrary roadblock procedure goes unproved. Edmond and LaFontaine involve different constitutional burdens, and compliance with such burdens is proved by different evidence. Neither Edmond nor LaFontaine, nor both read together, creates a new evidentiary “constitutional prerequisite” that a supervisor’s purpose for a roadblock program must be proved in order to prove the purpose of a roadblock.
Baker, 252 Ga. App. at 708-709 (Eldridge, J., dissenting) (emphasis in original).
We are not suggesting that it is improper for supervisory personnel to he present at a police checkpoint. To the contrary, having an experienced supervisor present may help ensure that proper procedures (including constitutional requirements) are followed by the officers conducting the roadblock. Nor is a checkpoint rendered unconstitutional solely because the supervisor who authorized the checkpoint later participates to some extent in the checkpoint’s operation. See, e.g., Johnson v. State, 320 Ga. App. 231, 233 (739 SE2d 718) (2013); Owens v. State, 308 Ga. App. 374, 376-377 (707 SE2d 584) (2011). This should be done with caution, however, as it can raise questions about whether the decision to implement the checkpoint was really made by the supervisor in advance rather than as an officer out in the field - particularly where, as here, the supervisor did not document when the decision to implement the checkpoint was made.
We also note in this respect that the majority below was wrong in saying that “officers are not precluded as a matter of law from acting simultaneously as a supervisor and a field officer,” to the extent that the majority was speaking to the decision to implement a checkpoint. Brown, 315 Ga. App. at 159 (emphasis added). As explained in Division 2 (b) above, the implementation decision must he made in advance by a supervisor acting in that supervisory capacity, not by an officer out in the field who happens also to have supervisory rank.