Field Sobriety Tests
Field sobriety tests in NC are three roadside exercises — the horizontal gaze nystagmus (HGN) eye test, the walk-and-turn and the one-leg stand — that officers use to decide whether to arrest you for DWI. They are scored by counting specific “clues,” and they are only as reliable as the officer’s instructions, the conditions on the roadside and the video that does (or doesn’t) back up the report.
What are standardized field sobriety tests in North Carolina?
Standardized field sobriety tests (SFSTs) are a battery of three divided-attention and eye tests developed with funding from the National Highway Traffic Safety Administration (NHTSA). North Carolina officers are trained on the NHTSA curriculum, and when a report says a driver “performed SFSTs,” it means the HGN, the walk-and-turn and the one-leg stand.
They are not the same as the “alphabet,” finger-to-nose or counting exercises some officers still use. Those are non-standardized tests: no validated scoring and no research-based decision point. They can still come into evidence as an officer’s observations, but they don’t carry the “validated” label prosecutors like to attach to the three standardized tests.
Field sobriety tests are one piece of the State’s case under G.S. 20-138.1. They don’t measure alcohol concentration. Their legal job is to help an officer decide whether there is probable cause to arrest. At trial, the State uses them as evidence that your mental or physical faculties were “appreciably impaired.” The roadside portable breath test and the Intox EC/IR II at the station are separate steps with separate rules.
Why “standardized” matters: the research behind the battery
“Standardized” means the officer must give the same instructions, look for the same clues and use the same scoring cutoff every time. The research numbers prosecutors cite only apply when the test is given and scored the way the research did it.
- 1977NHTSA-funded laboratory research reports that a combined score on three tests sorted subjects above or below 0.10 in 83% of cases.
- Early 1980sNHTSA research standardizes the battery. At the then-common 0.10 limit, about 77% of HGN classifications were correct, 68% for walk-and-turn and 65% for one-leg stand.
- Mid-1990sField studies in Colorado and Florida test the battery on real roadside stops.
- 1998NHTSA’s San Diego study (297 drivers, seven officers) validates the battery at 0.08 and adds the 4-clue HGN cutoff for 0.08.
- 2006North Carolina adds Rule 702(a1), letting trained officers testify to HGN results on the issue of impairment.
- 2017State v. Godwin: the NC Supreme Court holds a trained officer can be implicitly qualified to give HGN testimony.
NHTSA’s own training materials make the point directly: the validation applies when the tests are given in the prescribed, standardized way, scored with the standardized clues and judged by the standardized criteria. Change the instructions, skip a step or shorten the timing, and the research no longer describes the test you took. That gives the defense a lot to work with. You can read the current curriculum in NHTSA’s 2023 SFST participant manual.
The three field sobriety tests, clue by clue
Each test has a fixed list of clues and a “decision point,” the number of clues at which the officer is trained to predict an alcohol concentration of 0.08 or higher. Tap each tab for the instructions, the clues and the most common administration errors.
What the officer does: asks you to remove glasses, stand with feet together and follow a pen or penlight with your eyes only. The stimulus is held about 12 to 15 inches from your nose, slightly above eye level. Before scoring, the officer should check for equal pupil size, equal tracking and resting nystagmus, which are screens for medical problems.
The 6 clues (3 per eye):
- Lack of smooth pursuit: the eye jerks instead of gliding as it follows the stimulus (about 2 seconds out, 2 seconds back per pass).
- Distinct and sustained nystagmus at maximum deviation: obvious jerking with the eye held fully to the side for at least 4 seconds.
- Onset of nystagmus prior to 45 degrees: jerking starts before the eye reaches 45 degrees from center.
Decision point: 4 of 6 clues.
Common errors: passes done too fast, stimulus too close or too low, flashing patrol lights or passing traffic in the field of view, no medical screening, no vertical nystagmus check.
Instruction stage: you stand heel-to-toe on a real or imaginary line, left foot on the line and right foot in front, arms at your sides, while the officer explains and demonstrates the test.
Walking stage: nine heel-to-toe steps, a turn (keeping the front foot on the line and taking several small steps with the other), and nine heel-to-toe steps back, counting out loud and watching your feet.
The 8 clues:
- Can’t keep balance during the instructions
- Starts too soon
- Stops while walking
- Misses heel-to-toe (a gap of about half an inch or more)
- Steps off the line
- Uses arms to balance (raised about 6 inches or more)
- Improper turn
- Wrong number of steps
Decision point: 2 of 8 clues.
Common errors: sloped or gravel shoulder, no real line, bad footwear, instructions given while the person is already struggling to hold the starting position, no demonstration of the turn.
What the officer does: asks you to stand with feet together and arms at your sides, then raise either foot about 6 inches off the ground, keep both legs straight, look at the raised foot and count “one thousand one, one thousand two…” until told to stop. The officer times 30 seconds.
The 4 clues:
- Sways while balancing
- Uses arms to balance (about 6 inches or more)
- Hops
- Puts the foot down
Decision point: 2 of 4 clues.
Common errors: not timing the full 30 seconds (or letting the count run long), uneven surface, wind and traffic, high heels or boots left on, no questions about leg, back or inner-ear problems.
| Test | What it measures | Clues possible | Decision point | Accuracy at 0.08 (San Diego 1998) |
|---|---|---|---|---|
| Horizontal gaze nystagmus | Involuntary eye jerking | 6 | 4 | 88% |
| Walk-and-turn | Divided attention, balance, following instructions | 8 | 2 | 79% |
| One-leg stand | Divided attention, balance | 4 | 2 | 83% |
Score it yourself: the clue-counting quiz
Clue counting sounds mechanical, but small scoring calls decide whether a test is “failed.” Try these examples. They are hypothetical, except for the last one, which uses the scores reported in a real NC Supreme Court case.
Walk-and-turn (hypothetical): the driver takes 9 steps out and 10 steps back, and once raises an arm about 8 inches. How many clues?
- 1
- 2
- 3
Two: wrong number of steps, and using arms for balance. That is the 2-clue decision point, so the officer will record a “fail” even if every step was heel-to-toe.
HGN (hypothetical): lack of smooth pursuit in both eyes, plus distinct and sustained nystagmus at maximum deviation in the left eye only. Total?
- 2
- 3
- 4
Three clues of six, one short of the 4-clue decision point. Clues are scored per eye, which is why the order and timing of each pass matter.
One-leg stand (hypothetical): the driver sways, puts the foot down once at “one thousand eighteen,” then finishes. How many clues?
- 1
- 2
- 3
Two (sways, puts foot down). That meets the 2-clue decision point. The video should show whether the sway was real and whether the surface was level.
The driver can’t hold the heel-to-toe stance while the officer explains the walk-and-turn, and starts walking before being told. Clues so far?
- 2, before a single step is scored
- 1
- 0, because the walking stage hasn’t started
Two. Both are instruction-stage clues. That’s why lawyers check how long the officer made the person stand in that awkward position while talking.
In State v. Parisi (2019), the officer recorded 6 HGN clues, 1 walk-and-turn clue and 2 one-leg stand clues. Which tests reached their decision points?
- Only HGN
- HGN and one-leg stand
- All three
HGN (6 of 6, cutoff 4) and one-leg stand (2 of 4, cutoff 2). The walk-and-turn (1 of 8) did not. The Supreme Court still found probable cause based on all the facts together.
How accurate are field sobriety tests? What the studies do and don’t prove
In NHTSA’s 1998 San Diego study, officers using the full battery correctly classified drivers above or below 0.08 in 91% of cases. That number comes from one study with seven officers, and it measured something narrower than prosecutors sometimes suggest.
Keep three limits in mind. First, the studies measure whether the tests predict an alcohol concentration, not whether a person was actually “appreciably impaired,” which is what North Carolina law asks. Second, an 83% or 79% accuracy rate means a meaningful share of people tested were misclassified. Some of those were sober people who “failed.” Third, the research subjects were tested by officers who knew they were in a study, usually under better conditions than a dark shoulder of I-77 at 2 a.m.
None of that makes the tests worthless. It makes them evidence to be tested, not a verdict.
A field sobriety test is a measurement taken on the side of a road. Like any measurement, it is only as good as the conditions, the instructions and the person scoring it.Adkins & Purkey
Who shouldn’t be judged by these tests: injuries, age, weight, shoes and surfaces
The balance tests assume a healthy, steady person on a flat, dry surface. NHTSA’s training materials acknowledge that people over 65, people 50 or more pounds overweight and people with back, leg or inner-ear problems may have difficulty with the walk-and-turn and one-leg stand. They also tell officers to offer the chance to remove high heels.
- Medical conditions: knee and ankle injuries, back surgery, neuropathy, vertigo and inner-ear disorders, some eye conditions, and fatigue. (We’ve written up how an ankle injury can change the meaning of a walk-and-turn.)
- Medications: some lawful prescriptions affect balance or the eyes. See our guide to prescription medication and DWI.
- Surface and weather: slope, gravel, wet grass, wind, cold and the blast of passing trucks.
- Lighting and distractions: flashing emergency lights can affect the HGN test, and traffic noise affects whether instructions were heard.
- Footwear and clothing: heels, boots, sandals and a heavy coat on a cold night.
Do you have to take field sobriety tests in NC?
No North Carolina statute requires you to perform field sobriety tests, and the implied-consent penalties apply to chemical tests, not roadside exercises. But refusing isn’t consequence-free. Under G.S. 20-139.1(f), if you are charged with an implied-consent offense, evidence that you refused field sobriety tests the officer requested is admissible against you.
So the real question is not “Can I say no?” but “What will each choice look like in court?” A refusal can be argued as consciousness of guilt. A performance gives the officer clue counts. Neither choice is right for everyone. Declining field sobriety tests is also completely different from refusing the breath test at the station, which carries a 12-month license revocation. Our guide to willful refusal of a chemical test explains that separate rule.
How NC courts treat field sobriety test evidence: Rule 702(a1)
North Carolina treats HGN differently from the two balance tests. Walk-and-turn and one-leg stand are observations any witness can describe. HGN is treated as scientific evidence that requires a witness trained in it, and that witness may speak only to impairment, never to a specific alcohol concentration.
A witness may give expert testimony solely on the issue of impairment, and not on the issue of specific alcohol concentration level, about the results of an HGN test administered in accordance with the person’s training by someone who has successfully completed HGN training. A DHHS-certified Drug Recognition Expert may testify whether a person was under the influence of an impairing substance, and its category. (Close paraphrase of G.S. 8C-1, Rule 702(a1).)
Read Rule 702 on ncleg.gov ↗The path to that rule ran through the courts. In 1998 the Supreme Court held that HGN is scientific evidence and reversed a conviction where an officer testified about it without a proper foundation. The General Assembly answered in 2006 by adding subsection (a1). In 2011 it adopted the federal Daubert-style reliability test in Rule 702(a). Since then, the appellate courts have read (a1) as the legislature’s signal that HGN is reliable enough to admit, while holding the line on the “no numbers” limit. For a deeper dive, see our case note on State v. Godwin and our full guide to the HGN test.
HGN is a scientific test. Admitting an officer’s HGN testimony without evidence of reliability or judicial notice was prejudicial error, and the defendant got a new trial.
Read the opinion ↗An HGN-trained officer does not have to be formally tendered as an expert. Overruling the defense objection was an implicit finding that he was qualified. The Court read the 2006 amendment as a legislative signal that HGN results are reliable.
Read the opinion ↗At a Wilkesboro checkpoint, an odor of alcohol, glassy eyes, an admission to three beers and clues on all three SFSTs (6 HGN, 1 walk-and-turn, 2 one-leg stand) were enough for probable cause, even with no bad driving and no slurred speech.
Read the opinion ↗Weaving, odor, red glassy eyes, an admission and six HGN clues established probable cause. The trial court had found the officer shortened NHTSA’s HGN timing and treated that as going to weight, not admissibility.
Read the opinion ↗Odor of alcohol, bloodshot eyes, positive portable breath tests and signs of impairment on field sobriety tests at a checkpoint gave the officer probable cause.
Read the opinion ↗Physical evidence of impairment from sobriety tests (slurred speech, poor coordination) is not testimonial. A question requiring a testimonial answer, like the date of the driver’s sixth birthday, is.
Read the opinion ↗How field sobriety tests fit into probable cause
Field sobriety tests rarely win or lose a probable-cause fight alone. North Carolina courts look at the totality of the circumstances: the driving, odor, eyes, speech, admissions, test clues and anything that explains them.
Parisi and Woolard show how much weight appellate courts give clue counts when they’re combined with odor and an admission to drinking. Woolard is also a warning. An officer’s departures from NHTSA timing don’t automatically throw HGN out. They become a weight argument for the judge or jury. Compare State v. Overocker (2014), where a light odor and three drinks over four hours were not enough. Our Parisi case note walks through how these cases line up.
The defense lesson is practical. Probable cause can be challenged in a pretrial motion to suppress, and the officer’s description of each test is usually the heart of that hearing. If the video shows two clues where the report claims five, or a slope the report never mentions, the totality looks different.
How DWI lawyers review field sobriety test video
The best evidence of how a field sobriety test went is video. A careful review compares the dash-camera or body-camera footage, second by second, against the officer’s narrative and the NHTSA standards. Here is the checklist we work through. Our guide on how to review North Carolina DWI evidence covers the rest of the file.
- Lock down the footage early
Request dash-cam, body-cam and any jail or intake video quickly. Retention periods vary by agency, and footage that isn’t requested can be lost.
- Map the scene
Note the slope, surface, lighting, weather, traffic, emergency lights and the person’s shoes. Compare with what the report says, and with what it leaves out.
- Time the HGN
Count the seconds on each smooth-pursuit pass and each hold at maximum deviation. Check whether the medical screens and vertical check were done, and where the stimulus was held.
- Listen to the instructions
Were they complete, demonstrated and understood? Was the person left in the heel-to-toe stance for a long explanation? Did the officer ask about injuries?
- Re-score every clue
Score the walk-and-turn and one-leg stand independently, clue by clue, and compare against the report. Count the one-leg stand seconds.
- Look for what went right
Clear speech, steady exit from the car, normal walking and following directions are evidence too. The cases weigh the whole picture.
What to do if you were given field sobriety tests in NC
Write down everything you remember, while it’s fresh: where you stood, the surface, your shoes, the weather, what the officer said, what you said about injuries or medications, and roughly how long each test lasted. Then gather records of any medical conditions that affect balance or the eyes.
Don’t try to argue your performance with the officer or on social media. Bring your notes to a consultation. Our consultation checklist lists what to bring. Field sobriety evidence often shapes the bigger decisions in the case, including whether to file a suppression motion and whether to plead or go to trial. The full process is in our NC DWI guide.
From our book
Charged with DWI in North Carolina?
Our book walks through the roadside investigation from both sides of the badge, including how field sobriety tests are taught, given and challenged, so you can follow what happened at your stop.
Field sobriety tests NC: frequently asked questions
Do I have to take field sobriety tests in North Carolina?
No statute requires you to perform them, and they are not covered by the implied-consent penalties for chemical tests. But under G.S. 20-139.1(f), if you are charged with an implied-consent offense, evidence that you refused field sobriety tests the officer requested is admissible against you.
How many clues does it take to “fail” a field sobriety test?
Under NHTSA scoring, the decision points are 4 of 6 clues on the HGN, 2 of 8 on the walk-and-turn and 2 of 4 on the one-leg stand. Reaching a decision point is evidence, not proof of impairment.
Can the officer say my HGN test showed I was over 0.08?
No. Rule 702(a1) limits HGN testimony to the issue of impairment and bars testimony about a specific alcohol concentration. The Court of Appeals ordered a new trial in State v. Torrence (2016) when a deputy tied HGN clues to a probability of a high blood alcohol level.
Are field sobriety tests accurate?
In NHTSA’s 1998 San Diego study, the decision points matched a BAC at or above / below 0.08 in 88% (HGN), 83% (one-leg stand) and 79% (walk-and-turn) of cases, and 91% for the battery. That accuracy assumes standardized administration, and it still leaves room for misclassification.
What if I have a bad knee, back or inner ear?
Tell the officer, and tell your lawyer. NHTSA training materials acknowledge that people with back, leg or inner-ear problems, people over 65 and people 50 or more pounds overweight may have difficulty with the balance tests. Medical records can change how the results are viewed.
Adkins & Purkey, PLLC
Talk to a North Carolina DWI lawyer
If your arrest turned on field sobriety tests, we can get the video, re-score every clue and tell you what the evidence really shows. Offices in Huntersville (Lake Norman) and Locust (Stanly County), serving Charlotte, Mecklenburg, Iredell, Cabarrus, Stanly and courts across North Carolina.
General North Carolina legal information as of November 2026, not legal advice for any specific case. Reading this article does not create an attorney-client relationship. Attorney advertising.
