Skip to main content


Reset A A Font size: Print

Supreme Court of Georgia.



Decided: June 26, 2017

Appellant Quinton Mitchell appeals from the denial of his motion to suppress and motions in limine. Because the trial court failed to require the proper foundation for the Romberg field sobriety test under Harper v. State, 249 Ga. 519, 524-526 (1) (292 SE2d 389) (1982), we reverse on that ground. We affirm the remainder of the trial court's rulings.

When reviewing a trial court's ruling on a motion to suppress, an appellate court must construe the evidentiary record in the light most favorable to the factual findings and judgment of the trial court. This means that the reviewing court generally must accept the trial court's findings as to disputed facts unless they are clearly erroneous, although the reviewing court may also consider facts that definitively can be ascertained exclusively by reference to evidence that is uncontradicted and presents no questions of credibility, such as facts indisputably discernible from a videotape.

(Citations and punctuation omitted.) State v. Allen, 298 Ga. 1, 2 (1) (a) ( 779 SE2d 248) (2015).

So viewed, the facts show that Mitchell was stopped after a Fayette County sheriff's deputy noticed him failing to maintain his lane. When the deputy approached Mitchell, he smelled a strong odor of alcohol and noticed that he had difficulty removing his license from his wallet. Mitchell denied having had anything to drink, and when asked if he would perform field sobriety tests, he declined. He also declined to get out of his car. The deputy noted his slurred speech, and though Mitchell avoided eye contact, the deputy eventually saw that his eyes were “bloodshot and glassy.”

A Fayetteville Police Department officer arrived on the scene shortly afterwards, and the deputy told him that Mitchell had declined to perform field sobriety tests. The officer approached the car and had a short colloquy with Mitchell, who again refused to get out. The officer announced, “We got two ways we can do this. You can either get out of the car on your own, or we can get you out of the car. Okay. One way or another, you're coming out of the car.” Mitchell complied, and then stumbled and leaned against the car door for support. When Mitchell again refused to perform any field sobriety tests, the officer testified:

[I] told him that based on all the things that I observed already, which was the strong smell of alcohol coming from him; his mild slurred speech, his bloodshot, glassy eyes; the fact that he had to use the vehicle for ․ for balance, and his drunk-like appearance, his impaired appearance, that I believed that he was an impaired driver and that if he did not perform field sobriety, I had no option but to arrest him for DUI; or he could perform field sobriety and maybe he would, maybe he wouldn't; but that – there was no choice if not, so he submitted to field sobriety.1

Mitchell then agreed to perform the tests. After the officer and deputy stepped aside to discuss the matter, leaving Mitchell standing alone, the officer administered the tests. The officer testified that Mitchell “exhibited all six clues” on the HGN test, all eight clues in the walk and turn test, and declined to perform the one leg stand due to knee problems. Mitchell was outside the typical time estimation on the Romberg balance test, and during that test his eyes did not remain closed, he swayed back and forth, and his speech remained mildly slurred. The officer concluded that Mitchell was a less safe driver and so informed the deputy, who arrested Mitchell and read the implied consent warnings.

Mitchell was charged by accusation in the State Court of Fayette County with driving under the influence (less safe) and failure to maintain lane. He moved in limine to exclude, inter alia, the results of the field sobriety tests, and moved to suppress the evidence obtained by the police during the traffic stop. He also challenged the constitutionality of OCGA § 24-7-707, governing the admissibility of expert opinion testimony in criminal cases. At the hearing on Mitchell's motions, both officers testified, and the video of the traffic stop was played for the trial court. The trial court denied the motions and granted a certificate of immediate review. Mitchell filed an application for interlocutory appeal, which was granted by this Court.

When the evidence at a suppression hearing is uncontroverted and the credibility of witnesses is not in question, we conduct a de novo review of the trial court's application of the law to the undisputed facts. To the extent an issue concerns a mixed question of fact and law, we accept the trial court's findings on disputed facts and witness credibility unless they are clearly erroneous, but independently apply the law to the facts.

(Citations omitted.) Jones v. State, 291 Ga. 35, 36-37 (1) (727 SE2d 456) (2012). And we consider the officer's conduct of the traffic stop in its entirety. Cf. State v. Allen, supra, 298 Ga. at 1.

1. Mitchell asserts that the trial court erred in holding that the Romberg field sobriety test is not a scientific test requiring that a foundation be laid under Harper. We agree.

The Harper decision guides a trial court's determination of whether a scientific principle or technique is competent evidence in a criminal case:

It is proper for the trial judge to decide whether the procedure or technique in question has reached a scientific stage of verifiable certainty, or in the words of Professor Irving Younger, whether the procedure “rests upon the laws of nature.” The trial court may make this determination from evidence presented to it at trial by the parties; in this regard expert testimony may be of value. Or the trial court may base its determination on exhibits, treatises or the rationale of cases in other jurisdictions. The significant point is that the trial court makes this determination based on the evidence available to him rather than by simply calculating the consensus in the scientific community.

(Citations and footnote omitted.) 249 Ga. at 525-526 (1). But a threshold issue is presented in determining whether the procedure or technique “deal[s] with scientific principles [or] observation and comparison of physical objects, with matters not of science but of skill and experience.” Belton v. State, 270 Ga. 671, 673-674 (4) (512 SE2d 614) (1999).2

In Belton, this Court looked to Court of Appeals decisions on field sobriety tests to determine, by analogy, that the comparison of shoe prints with shoes is not “a matter of scientific principle or technique.” Id. at 674 (4). Similarly, in determining the applicability of the Harper analysis to field sobriety tests, the Court of Appeals has considered whether the principles or techniques in question are properly a subject of scientific analysis under Harper, or are merely well-known consequences of intoxication, “as obvious to the layperson as to the expert.” Hawkins v. State, 223 Ga. App. 34, 36 (1) (476 SE2d 803) (1996). We must therefore determine whether the Romberg test falls into the category of a simple physical dexterity exercise observable by the average layperson, such as the walk and turn test or the one-leg stand test, or a scientific test which must meet standards of validity and reliability, such as the horizontal gaze nystagmus or “HGN” test. See generally Stewart v. State, 280 Ga. App. 366, 368-369 (2) (634 SE2d 141) (2006); State v. Pastorini, 222 Ga. App. 316, 318-319 (2) (474 SE2d 122) (1996).3

In the case before us, the only witnesses at the hearing on the motions in limine were the police officers, and only the city police officer testified with respect to “the Romberg balance [sic.].” He testified that the subject is instructed to shut his eyes, tilt his head backwards, and estimate the passage of 30 seconds. He added that “[t]he main purpose of [the test] is to get a person's – excuse me – to gauge a person's internal clock, to figure out if their internal clock is correct or accurate, given that certain drugs, alcohol being one of them, that could impair your ability to interpret the passage of time or perceive it.” The officer elaborated that he was also looking for “eyelid tremors” and that “pretty much anybody that's unimpaired and of reasonable faculty can estimate, can get within five seconds, plus or minus, of that 30-second mark.”

On cross-examination, however, the officer acknowledged that he was not aware of any validation studies for the Romberg test, “not like there are for the other three tests, no, sir.” He also acknowledged that the range of plus or minus five seconds as an indication of impairment had not been established. His knowledge of the test was based on his participation in DRE or “drug recognition expert” school. No scientific or medical testimony was presented at the hearing.4

We conclude that, on the basis of the evidence presented at the hearing in this case, admissibility of the Romberg test is subject to the Harper standard. The significance of eyelid tremors or an individual's “internal clock,” how they may be affected by the consumption of alcohol, and particularly whether a range of five seconds above or below the actual passage of 30 seconds establishes impairment, are not matters of common sense or experience, nor are they obvious to the average lay observer.5 The trial court therefore erred in failing to conduct a Harper analysis, whether through the evaluation of expert testimony or through the examination of exhibits, treatises, or the law of other jurisdictions. Harper, supra, 249 Ga. at 525-526 (1). We therefore reverse this portion of the trial court's order.

2. Mitchell also contends the trial court erred in denying his motion to suppress the results of the field sobriety tests, because he was told that he would be arrested if he did not submit to the tests, but was not given Miranda warnings.

[U]nder Georgia law Miranda warnings must precede a request to perform a field sobriety test only when the suspect is “in custody.” The test of “in custody” is whether a “reasonable person in the suspect's position would have thought the detention would not be temporary.”

(Citations and footnotes omitted.) Price v. State, 269 Ga. 222, 225 (3) (498 SE2d 262) (1998). In Price, this Court held that failure to give Miranda warnings rendered evidence of field sobriety tests inadmissible. There, the officer told appellant that her license was suspended, that he believed she was intoxicated based on his observations of her behavior, and that “he would take her to jail for DUI regardless of whether she performed the field evaluations.” Id. We held that “having been informed that she was going to jail, a reasonable person would have believed that the detention was not temporary.” Id. Similarly, in Hale v. State, 310 Ga. App. 363 (714 SE2d 19) (2011), once the officer “replied in the affirmative” when appellant asked if he was going to jail, “a reasonable person in [appellant's] position would believe that his or her freedom of action had been more than temporarily curtailed, which thereby placed him in custody for purposes of Miranda.” Id. at 365-366 (1). See also State v. Norris, 281 Ga. App. 193, 196 (635 SE2d 810) (2006) (reasonable person would believe he was under arrest when officer told him to turn around and put his hands behind his back).

In contrast, “absent the officer making any statement that would cause reasonable person to believe that [ ]he was under arrest and not [merely] temporarily detained during an investigation, the officer's belief that probable cause exists to make an arrest does not determine when the arrest is effectuated until the officer overtly acts so that a reasonable person would believe [ ]he was under arrest.” State v. Kirbabas, 232 Ga. App. 474, 476 (502 SE2d 314) (1998). Here, the officer did not indicate by words or actions that he was going to arrest Mitchell. Rather, he informed Mitchell that he was not yet under arrest, and if “he could perform field sobriety ․ maybe he would, maybe he wouldn't” be arrested. As in Rowell v. State, 312 Ga. App. 559, 565 (718 SE2d 890) (2011), the officer “gave [Mitchell] an option of sorts: perform the test properly or go to jail. Thus a reasonable person in [Mitchell]'s position would have believed that [ ]he was not yet under arrest and that [his] detention still could be only temporary.” And, as in State v. Mosley, 321 Ga. App. 236 (739 SE2d 106) (2013), Mitchell was informed that he was not under arrest, was not placed in handcuffs, and was left alone when the officer stepped aside to consult with the sheriff's deputy, who was “conducting field-sobriety testing for the very purpose of determining whether to take [Mitchell] into custody.” Id. at 239. A reasonable person therefore would not have believed that he was in custody, and the trial court did not err in denying Mitchell's motion to suppress on this ground.

3. Mitchell asserts that the trial court erred in failing to exclude evidence of his initial refusal to submit to field sobriety tests, because such tests ought to be treated as a warrantless search, analogous to the warrantless search of the appellant's vehicle in Mackey v. State, 234 Ga. App. 554, 555-556 (507 SE2d 482) (1998). Mitchell argues from Mackey that “an individual should be able to invoke his Fourth Amendment rights without having his refusal used against him at trial.” 234 Ga. App. at 556. But the Court of Appeals in Mackey observed, without further elaboration, that “[a] defendant's refusal to consent to a warrantless search of his vehicle or other property is quite a different issue” from “a defendant's refusal to submit to a blood or urine test for determining alcohol or drug content.” Id. at 555-556.

The State, on the other hand, argues that the Court of Appeals more directly considered the issue in Long v. State, 271 Ga. App. 565, 568-569 (2) (610 SE2d 74) (2004). After noting that this Court has held that evidence of pre-arrest refusal to undergo a field sobriety test does not violate a defendant's right against self-incrimination, citing Keenan v. State, 263 Ga. 569, 572 (2) (436 SE2d 475) (1993), the Court of Appeals rejected Long's contention that “commenting on her refusal to submit to the field tests is equivalent to commenting on the failure to consent to a warrantless search.” 271 Ga. App. at 568 (2). We agree with the Court of Appeals' rejection of this argument, but for the following reasons.6

Generally, in a “search” of an individual, some tangible evidence is taken from that person: whether a physical object in the person's possession, or a sample of some part of their body, such as hair, blood, or urine. An action by the State which does not obtain any tangible item, but merely obtains information as to “personal characteristics,” lies in a middle ground. See generally Wayne R. LaFave, Search and Seizure § 2.6 (a) (5th ed. 2016). The United States Supreme Court has concluded that, on the one hand, “searches” include such actions as taking blood, breath, or urine samples, Skinner v. Railway Labor Executives' Assn., 489 U. S. 602, 616-618 (II) (B) (109 SCt 1402, 103 LE2d 639) (1989),7 removing scrapings from underneath an individual's fingernails, Cupp v. Murphy, 412 U. S. 291, 295 (93 SCt 2000, 36 LE2d 900) (1973), or obtaining DNA evidence via a cheek swab. Maryland v. King, __ U. S. __ (133 SCt 1958, 186 LE2d 1) (2013). But the high court has also held that observation of actions performed by an individual is not a search, even if performed at the demand of the State. See, e.g., United States v. Mara, 410 U. S. 19, 21 (93 SCt 774, 35 LE2d 99) (1973) (handwriting exemplar); United States v. Dionisio, 410 U. S. 1, 14 (93 SCt 764, 35 LE2d 67) (1973) (voice exemplar). In Cupp, supra, the high court, citing Davis v. Mississippi, 394 U. S. 721, 726-727 (89 SCt 1394, 22 LE2d 676) (1969), observed that fingerprinting is a device to obtain a record of a “mere physical characteristic,” analogous to voice or handwriting. 412 U. S. at 295. A field sobriety test appears to us to be an act more akin to a handwriting or voice exemplar than the physical removal of tangible evidence.

In his supplemental brief, Mitchell points to decisions of other state courts holding that a field sobriety test is a “search” within the meaning of the Fourth Amendment, notably Blasi v. State, 167 Md. App. 483, 504-505 (893 A2d 1152) (2006), Ackerman v. State, 774 NE2d 970, 980 (2) (a) (Ind. App. 2002); Hulse v. Dept. of Justice, 289 Mont. 1, 30 (961 P2d 75) (1998), and State v. Nagel, 320 Ore. 24, 35 (880 P2d 451) (1994).8 These courts base their holdings on the concept that field sobriety tests elicit physical activities that are not “the sort of normal physical activity that one displays to the public,” Blasi, supra, at 504-505, and moreover may reveal private information not ordinarily visible to the public, such as that the subject “is illiterate, has Alzheimer's disease, or suffers from multiple sclerosis.” Id. at 503; see also Hulse, supra, 289 Mont. at 30; Nagel, supra, 320 Ore. at 35-36.

But the fact that a test may incidentally reveal some other condition or impairment does not necessarily render it a search within the meaning of the Fourth Amendment. A handwriting exemplar, for example, while not a search, see Mara, supra, would certainly reveal the subject's illiteracy and might also reveal a neurological condition. And while field sobriety tests may involve specific, unusual maneuvers that are not normally performed in public or private, most are simply intended to reveal, more quickly and in a reproducible fashion, matters that also would be revealed by more time-consuming, but clearly permissible, passive observation.9 Such characteristics as unsteady gait, lack of balance and coordination, impaired speech, lack of memory, or inability to divide one's attention, generally would become apparent to a casual observer over a longer period of time.

Although it is a close question, we conclude that a basic field sobriety test is not a search implicating Fourth Amendment protections.10 Mitchell's Fourth Amendment protections were not implicated, and we decline to apply the rule enunciated in Mackey to the administration of field sobriety tests. The trial court accordingly did not err in denying Mitchell's motion to suppress his refusal to submit to field sobriety testing.

4. Finally, Mitchell contends the trial court erred in rejecting his equal protection and separation of powers challenges to OCGA § 24-7-707, governing expert testimony in criminal proceedings. As the trial court noted, however, this Court rejected the equal protection argument with regard to the distinct provisions governing expert testimony in civil and criminal proceedings in Mason v. Home Depot USA Inc., 283 Ga. 271, 276-277 (2) and (3) (658 SE2d 603) (2008) (construing former OCGA § 24-9-67 11 and OCGA § 24-9-67.1 12 ); see also Zarate-Martinez v. Echemendia, 299 Ga. 301, 304 (2) n.2 (788 SE2d 405) (2016). In the absence of any argument that we overrule Mason, we decline to do so.13

Mitchell's assertion that OCGA § 24-7-707 violates the separation of powers clause of the Georgia Constitution by usurping the judicial power to determine the admissibility of evidence is also without merit. In Zarate-Martinez, supra, this Court rejected a similar argument with respect to OCGA § 24-7-702 (c), governing expert witnesses in professional malpractice actions:

In direct contradiction to [appellant's] argument, the Georgia Constitution specifically provides that “[a]ll rules of evidence shall be as prescribed by law.” Ga. Const. of 1983 Art. VI, Sec. I, Par. IX. By providing evidentiary guidance to the judiciary through the passage of OCGA § 24-7-702 (c), the General Assembly has simply acted consistently with its constitutional duty, rather than in contravention of it. See Bell v. Austin, 278 Ga. 844, 846 (2) (607 SE2d 569) (2005) (“[T]he legislature has power to establish rules of evidence where not in conflict with the constitution or rights guaranteed by it”) (citation and punctuation omitted).

299 Ga. at 310 (2) (e). Similarly, the language in OCGA § 24-7-707, which has been the law of Georgia for over 150 years,14 provides a statutory framework within which the trial court determines the qualifications of the expert and the admissibility of the proffered testimony. This Court's rationale in Zarate-Martinez applies, OCGA § 24-7-707 does not violate the separation of powers clause, and Mitchell's argument therefore fails.

Judgment affirmed in part and reversed in part. All the Justices concur.


1.   The State asserts that the officer was mistaken when he testified at the hearing, and that we should conclude from the video that the officer did not assert that he would arrest Mitchell unless he performed the field sobriety tests. The dashcam video recording shows that the officer said, “It would behoove you to cooperate with me, you know that, right, but if you don't ․” before Mitchell interrupted him. But the video also shows that the conversation continued after the interruption, although it became inaudible. We therefore cannot conclude that it is “indisputably discernible,” State v. Allen, 298 Ga. 1, 2 (1) (a) (779 SE2d 248) (2015), that the video recording contradicts the officer's explicit and detailed testimony at the hearing. And the trial court summarized the officer's hearing testimony in the “factual background” of its order on the motion to suppress, thereby in effect giving it credence.

2.   Although Georgia's new Evidence Code is applicable to the trial of this case, the evidentiary requirements relating to the admissibility of expert opinion testimony in a criminal case under the new Evidence Code (OCGA § 24-7-707) are nearly identical to those that applied under the former Evidence Code (OCGA § 24-9-67). Accordingly, it is appropriate to rely, as we do in this case, on decisions under the old Code. See Jones v. State, 299 Ga. 40, 42 (2) n. 2 (785 SE2d 886) (2016).Mosby v. State, 300 Ga. 450, 453 (2) n. 2 (796 SE2d 277) (2017).

3.   The Court of Appeals has noted, but not evaluated, the Romberg test in several opinions. In Kar v. State, 318 Ga. App. 379 (733 SE2d 387) (2012), it posed but did not reach the question of whether the Romberg test is subject to the Harper requirements, in light of the overwhelming evidence of the defendant's guilt. Id. at 381 (1).

4.   In its brief on appeal, the State cites a medical journal article for the history of the Romberg test. But the State fails to note that, while the authors list over 20 medical conditions that can result in a positive Romberg's sign other than its “hallmark” as a symptom of sensory ataxia, including neurosyphilis (the “classical cause”) and alcohol consumption predisposing to vitamin B12 deficiency, they do not mention intoxication or field sobriety as a proper subject for the test. Nor do they mention counting any elapsed period of time as part of the Romberg test. See A. Khasnis and R. M. Gokula, Romberg's Test, JOURNAL OF POSTGRADUATE MEDICINE, Vol. 49, No. 2, April-June 2003, pp. 169-172, available at

5.   Courts of the various states which have considered the numerous variations of the Romberg test do not agree on its name, its elements, its significance, its methodology, or even whether it is a recognized field sobriety test. See, e.g., Gradford v. Huntsville, 557 So2d 1330, 1331 (Ala. Crim. App. 1989) (appellant failed “Romberg alphabet test”); Peoplev. Carlson, 677 P2d 310, 316 n.6 (Colo. 1984) (“The Rhomberg” [sic] requires the subject to stand with hands at sides, feet together, eyes closed and head tilted back, for twenty seconds; “[p]ersons under the influence will sway from side to side or front to rear. Persons not under the influence will move in a circular-type motion.”); Commonwealth v. Jones, 88 Mass. App. 1120 n.1 (43 NE3d 349) (2016) (unpublished opinion) (trooper testified that “defendant exhibited pronounced swaying and eyelid tremors indicative of marijuana that is psychoactive in the system” while performing “modified Romberg balance test”); State v.Klawitter, 518 NW2d 577, 579 (Minn. 1994) (“Romberg test” used in drug recognition protocol but not in field sobriety testing); Commonwealth v. Keppel, 2016 Pa. Super. Unpub. LEXIS 2600 (2016) (trooper testified, “The Romberg Balance Test is a test designed to see if the defendant is under the influence of any drugs”); Buchanan v. State, 1990 Tex. App. LEXIS 991 (1990) (officer testified that Romberg test required subject to “sit in an erect position with the head tilted back and eyes closed. The officer looks for swaying.”); Johnsonv. State, 2009 Tex. App. LEXIS 3871 n. 1 (2009) (police officer testified that “the Romberg balance test was not a standardized field sobriety test, but was nonetheless ‘a test that I give everyone.’ ”).

6.   We express no opinion as to the State's argument or the Court of Appeals' reliance on Keenan.

7.   The Skinner court noted that the testing of urine, while not as obviously invasive of the person as a blood or breath test, is nevertheless a search because of the personal and private nature of the excretory functions. 489 U. S. at 617 (II) (B).

8.   People v. Walter, 374 Ill.App.3d 763, 772-773 (2007), cited by Mitchell, declines to reach the issue.

9.   We note that “field sobriety test” is a somewhat loosely defined term, as the tests involved may range from simple physical acts such as touching a finger to the nose or walking a straight line, through tasks such as reciting the alphabet or counting, to more unusual activities such as the HGN or Romberg tests.

10.   Although the courts of some of our fellow states have reached a different result, the overwhelming majority conclude that, even though a field sobriety test constitutes a search, it may be performed on the basis of a reasonable suspicion that the subject was driving under the influence. See Blasi, supra, 167 Md. App. at 511; Hulse, supra, 289 Mont. at 37; McCormick v. Municipality of Anchorage, 999 P2d 155, 160 (Alaska App. 2000); Statev. Ferreira, 133 Idaho 474, 480-481 (II) (A) (988 P2d 700) (Idaho App. 1999); State v.Taylor, 648 So2d 701, 703-704 (II) (Fla. 1995); State v. Lamme, 19 Conn. App. 594, 599-600 (563 A2d 1372) (1989), aff'd, 216 Conn. 172 (579 A2d 484) (1990); State v. Gray, 150 Vt. 184, 191 (III) (552 A2d 1190) (1988); Dixon v. State, 103 Nev. 272, 273 (737 P2d 1162) (1987); State v. Superior Court, 149 Ariz. 269 (1) (718 P2d 171) (1986); State v. Stevens, 394 NW2d 388, 391 (Iowa 1986), cert. denied, 479 U. S. 1057 (93 LE2d 986, 107 SCt 935) (1987); State v. Wyatt, 67 Haw. 293, 305 (III) (B) (687 P2d 544) (1984); State v. Little, 468 A2d 615, 617 (Me. 1983); People v. Bennett, 189 Cal. Rptr. 77, 80 (Cal. App. 1983). While the courts of Colorado and Oregon have held that probable cause is required to administer a field sobriety test, see People v. Carlson, 677 P2d 310, 316-317 (III) (Colo. 1984), and Nagel, supra, 320 Ore. at 34-37, they stand alone in so holding. But we need not reach this issue.

11.   Now OCGA § 24-7-707: “In criminal proceedings, the opinions of experts on any question of science, skill, trade, or like questions shall always be admissible; and such opinions may be given on the facts as proved by other witnesses.”

12.   Now OCGA § 24-7-702, applicable to civil actions.

13.   Mitchell contends that his case presents “a unique scenario” because he asserts – without further elaboration – that “the same acts that provide the basis for a criminal charge may also provide the basis for a civil suit.” But this is true in any criminal prosecution involving an act against a victim that may also form the basis for a cause of action in tort.

14.   See former Code 1863 § 3792; former Code 1873 § 3868; former Code 1882 § 3868; former Civil Code 1895 § 5287; former Penal Code 1895 § 1022; former Civil Code 1910 § 5876; former Penal Code 1910 § 1048; former Code 1933 § 38-1710; former OCGA § 24-9-67.

BOGGS, Justice.

Copied to clipboard