The Modified Romberg, Finger-to-Nose and Other Non-Standardized Field Sobriety Tests in NC

Field sobriety tests · Beyond the big three

North Carolina officers sometimes ask drivers to do more than the three standardized field sobriety tests. They may ask you to tilt your head back, close your eyes and guess when 30 seconds have passed (the Modified Romberg balance), touch your nose with your eyes closed, count on your fingers, recite part of the alphabet or count backwards. These non-standardized tests look official, but NHTSA has not identified validated clues for any of them. Here is how each one works, how North Carolina courts treat them, and the questions a defense lawyer asks about them.

What is a non-standardized field sobriety test?

A non-standardized field sobriety test is any roadside coordination or attention exercise that is not one of the three tests in NHTSA’s Standardized Field Sobriety Test (SFST) battery. The standardized battery is horizontal gaze nystagmus (HGN), the walk-and-turn and the one-leg stand. NHTSA’s DWI Detection and SFST Refresher manual (revised 10/2015) describes those three as “the only Standardized Field Sobriety Tests for which validated clues have been identified.”

Everything else is extra. The extras most often seen in North Carolina police reports and videos are the Modified Romberg balance, finger-to-nose, finger count, the alphabet test and counting backwards. None has a NHTSA list of validated clues or a decision point. That does not make them meaningless, and it does not make them inadmissible. It means they carry less weight, and an officer who says a driver “failed” one is describing a personal judgment, not a scored result. For the basics of the big three, see our guide to standardized field sobriety tests in North Carolina and our deep dive on the HGN test.

The Modified Romberg balance test, step by step

The Modified Romberg asks a person to stand still with the eyes closed and estimate the passage of 30 seconds. It borrows its name from the Romberg test doctors use in neurological exams, which also checks balance with the eyes closed. It is one of the four “psychophysical” tests in step 5 of the Drug Recognition Expert (DRE) evaluation, according to the International Association of Chiefs of Police’s description of the 12-step DRE process.

Detailed U.S. DRE manuals are not freely published, but the United Kingdom’s Code of Practice for Preliminary Impairment Tests (Department for Transport, April 1, 2017) publishes a full version of the test. In that version:

1. Starting position

   Heels and toes together, arms down at the sides, while the officer gives instructions.

2. Head back, eyes closed

   On the officer’s signal, the person tilts the head back slightly and closes the eyes.

3. The internal clock

   The person stays in that position and, when they think 30 seconds have passed, brings the head forward, opens the eyes and says “stop.”

4. What gets written down

   Whether the person kept balance during instructions; whether they stepped, swayed or raised their arms; whether they opened their eyes or lowered their head early; the number of seconds that actually elapsed; and whether they finished.

Sway. Officers usually describe sway as front-to-back or side-to-side and sometimes estimate it in inches. There is no published, validated cutoff for how much sway means impairment. Standing with the feet together and the eyes closed removes the visual cues people rely on to stay upright, so some sway is normal.

The 30-second estimate. The point of the time estimate is to see whether a person’s sense of time seems to be running fast or slow. But people vary, and nerves, fatigue and the stress of a roadside stop can all change how someone counts. A sober person who calls “stop” at 22 seconds or 38 seconds has not failed anything; there is no validated pass/fail range. The U.K. Code itself states that its impairment tests are not passed or failed and have no scoring system.

Eyelid tremors. In drug cases, DRE reports sometimes note eyelid tremors seen while the eyes are closed during the Romberg. The public procedure we could confirm (the U.K. Code) lists sway, stepping, arm movement, eye opening and the elapsed time, but not eyelid tremors. If a report relies on eyelid tremors, a defense lawyer will ask what training defines them, how they were observed in roadside lighting, and whether the video shows them.

Finger-to-nose, finger count, alphabet and counting tests

Finger-to-nose. In the version printed in the U.K. Code, the person stands with feet together, arms out in front, fists closed with the index fingers extended, then tilts the head back and closes the eyes. The officer calls out a hand in a set order (left, right, left, right, right, left), and the person touches the tip of the nose with the tip of that index finger. The officer records swaying or stepping, whether the correct hand was used, and any touch to a part of the face other than the tip of the nose. The IACP lists finger-to-nose as the fourth psychophysical test in a DRE evaluation.

Finger count. A common version has the person touch the thumb to each fingertip of one hand while counting “one, two, three, four,” then back “four, three, two, one,” for several rounds. Officers watch for miscounts, skipped fingers and starting or stopping too early.

Alphabet and counting. The officer asks the person to recite a stretch of the alphabet (for example, from J to V) without singing, or to count backwards between two numbers. These tests check attention and recall more than balance. They also involve speech, which raises a Fifth Amendment question discussed below.

Why these tests are not NHTSA-validated

NHTSA’s field sobriety research tested specific exercises, given with specific instructions and scored with specific clues, against measured alcohol concentrations. The standardized battery came out of that work. According to NHTSA’s refresher manual, the original Southern California Research Institute (SCRI) laboratory research (reported in 1977 and 1981) found HGN alone was 77% accurate, the walk-and-turn 68% and the one-leg stand 65%. In the 1998 San Diego field study by Stuster and Burns (DOT HS 808 839), the manual reports HGN at 88%, walk-and-turn at 79%, one-leg stand at 83%, and correct arrest decisions 91% of the time using the full battery at 0.08 and above.

The non-standardized tests never went through that process as roadside alcohol screens. They have no agreed instructions, no defined clues and no decision point. Two officers can give the “same” alphabet test with different letters, different instructions and different ideas of what a mistake looks like. That is the core reason defense lawyers push back when these tests are presented as if they were scientific.

The Romberg inside a DRE evaluation

The Modified Romberg and finger-to-nose matter most when a Drug Recognition Expert is involved. In the IACP’s 12-step process, step 5 is four divided-attention tests: the Modified Romberg balance, walk-and-turn, one-leg stand and finger-to-nose. A DRE combines those results with eye exams, vital signs, pupil size in three lighting conditions, muscle tone, an interview and a toxicology sample to form an opinion about whether the person is impaired by drugs and, if so, which drug category.

North Carolina gives certified DREs special status. Under Rule of Evidence 702(a1)(2), a witness who holds a current DRE certification from the Department of Health and Human Services may give expert testimony solely on whether a person was under the influence of an impairing substance and the category of that substance. A patrol officer without DRE certification has no such statutory shortcut for Romberg or finger-to-nose results. For drug cases more broadly, see our guides to prescription-drug DWI and medication and DWI.

How North Carolina officers use non-standardized tests

In practice, officers use these tests in three ways. First, as add-ons after the standardized battery, to build a fuller picture for probable cause. Second, as substitutes when a driver cannot fairly do the walk-and-turn or one-leg stand because of an injury, a medical condition, footwear or the roadside surface. Third, as part of a DRE evaluation after arrest.

The State does not need any field sobriety test to prove DWI. Under G.S. 20-138.1, the offense can be proved by showing a person drove while under the influence of an impairing substance (appreciable impairment), after having consumed enough alcohol to have an alcohol concentration of 0.08 or more at any relevant time after the driving, or with any amount of a Schedule I controlled substance or its metabolites in the blood or urine. “Impairing substance” is defined in G.S. 20-4.01. Field tests, standardized or not, are evidence the judge or jury weighs alongside driving behavior, odor, speech, statements and any chemical result. A preliminary roadside breath screen is a different tool; see our guide to the portable breath test.

How North Carolina courts treat these tests

North Carolina appellate courts have not singled out the Modified Romberg or alphabet test for a special admissibility rule. Instead, they treat performance on these tasks as one more observation in a totality-of-the-circumstances analysis, while drawing firm lines around expert testimony and around anything testimonial.

Observations count, and so do good performances. In Rawls v. Peters (1980), difficulty touching the nose was one of several observations (with a strong odor of alcohol, red and glassy eyes and staggering) that together supported probable cause, according to the UNC School of Government’s summary. In State v. Townsend (2014), the officer asked a checkpoint driver to recite the alphabet from J to V, and he did it without incident. The Court of Appeals still found probable cause based on other facts, including the odor of alcohol, bloodshot eyes, an admission to drinking, clues on the three standardized tests and two positive Alco-Sensor results. Read the other way, Townsend shows that a clean alphabet test goes into the record and should be argued; it does not disappear because other evidence pointed toward impairment.

Expert testimony is limited by Rule 702. In State v. Godwin (2017), the Supreme Court of North Carolina held that a trial court need not formally announce that an officer is an expert before the officer testifies about HGN; the court can implicitly find the officer qualified. In State v. Fincher (2018), the Court of Appeals held that a certified DRE could testify that a driver was impaired by a central nervous system depressant without a separate showing that the DRE protocol is reliable, because Rule 702(a1)(2) authorizes that testimony. In Fincher, the driver wore a boot on one leg, so officers gave HGN but not the other field tests before arrest. In State v. Moore (2025), however, the Court of Appeals held that a DRE must still reliably apply the method to the facts, and an opinion based on an after-the-fact review of records and the trial evidence, instead of an actual DRE evaluation, should have been excluded, though the error was harmless given the other evidence of guilt.

Physical tests are not testimonial, but some answers are. In State v. Flannery (1976), the Court of Appeals held that physical dexterity tests are “not evidence of a testimonial or communicative nature,” so Miranda warnings were not required before them. In Pennsylvania v. Muniz (1990), the U.S. Supreme Court treated slurred speech and poor coordination as physical evidence, but held that the driver’s answer to “the date of your sixth birthday” was a testimonial response that required Miranda warnings in custody. The Court did not decide whether counting during a test is testimonial. Courts in other states have split on alphabet and counting tasks, as the UNC School of Government has explained. Under Berkemer v. McCarty (1984), ordinary roadside questioning during a traffic stop usually is not custodial, which is why most Fifth Amendment arguments about these tests arise after arrest, such as during a DRE evaluation.

Key cases on non-standardized tests and officer testimony

Browse more decisions in our North Carolina DWI case law library, including the field sobriety test cases, and our case notes on State v. Godwin, State v. Parisi and State v. Townsend.

Cross-examination themes for non-standardized tests

A good cross-examination on these tests is calm and specific. The goal is not to call the officer dishonest; it is to show the jury or judge how little a non-standardized test can tell them. Common themes:

1. “It isn’t one of the three”

   Establish that the officer’s own SFST training identifies only HGN, walk-and-turn and one-leg stand as the tests with validated clues, and that the Romberg, finger-to-nose, finger count, alphabet and counting tests are not among them.

2. The exact instructions

   Ask what words were used, in what order, and whether there was a demonstration. With no standard script, unclear instructions are easy to give and hard to defend. Compare the testimony with the body-cam or dash-cam video.

3. No clues, no decision point

   Ask how many “mistakes” equal impairment. There is no published answer. An officer who testifies that a driver “failed” the alphabet test is offering an opinion without a standard.

4. The 30-second estimate

   Ask what range of estimates is normal for sober people, where that number comes from, and whether the officer timed the estimate with a watch or simply counted in his head.

5. Conditions and the person

   Explore the surface, slope, wind, traffic, flashing lights, footwear, cold, age, weight, injuries, inner-ear or neurological conditions, dexterity problems, language, education and anxiety. Each affects balance, counting or recall.

6. What went right

   Bring out every part of every test the driver did correctly, as in Townsend‘s clean alphabet recitation. Juries notice when only the misses were written down.

7. Training and certification

   Is the officer a currently certified DRE? If not, Rule 702(a1)(2) does not apply. If so, did the officer actually perform the evaluation, as Moore requires, or form the opinion some other way?

8. File motions on time

   Suppression issues in district court DWI cases generally must be raised before trial under G.S. 20-38.6.

For a broader checklist, see how to review North Carolina DWI evidence and our case study on field tests and an ankle injury.

Quiz: standardized or not?

Check what you learned. Each answer is based on the sources and cases discussed above.

When a test has no standard instructions and no validated clues, the officer’s description becomes the score. The video is how you check it.Adkins & Purkey

Top 10 FAQs: non-standardized field tests

Related guides

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. Past results do not guarantee a similar outcome. Attorney advertising.