The One-Leg Stand Test

There are only four clues on the one leg stand, and two of them fail you. That is half the scoring sheet.

Two of the nine things you are expected to do are not in the instructions at all. You are told to raise your foot, keep your legs straight, look at the foot, and count out loud. You are never told not to hop, and you are never told not to raise your arms. Both are scored against you anyway.

And there is a common officer error that manufactures a clue outright. It is described below, and it is worth checking the video for in every case.

How the Test Is Administered

The one-leg stand is a divided attention test. It asks you to balance, listen, and count at the same time. None of those is hard by itself. The combination is the point.

The test runs in two stages, and both are scored.

The instruction stage. You stand with your feet together and your arms down at your sides while the officer explains and demonstrates the test. The officer is supposed to tell you not to start until told to do so, and to confirm that you understand. You are already balancing during this stage, on a narrower base than you would normally choose, while trying to absorb a set of instructions you are hearing for the first time.

The balance and counting stage. On the officer’s word, you raise one foot, either foot, approximately six inches off the ground, keeping the raised foot parallel to the ground. You keep both legs straight and your arms at your sides. You look at the raised foot. And you count out loud in a specific way: one thousand and one, one thousand and two, one thousand and three, and so on, until told to stop. The officer is supposed to time it and discontinue the test at 30 seconds.

The officer has duties of his or her own here. The officer is instructed to demonstrate the test, to observe from a safe distance, and to minimize his or her own movement so as not to interfere. If you put the raised foot down, the officer is supposed to tell you to pick it back up and continue counting from where you left off. If you count very slowly, the officer terminates the test at 30 seconds. If you cannot do the test at all, the officer is supposed to record whatever clues were observed and document why the test was not completed.

The test also has a surface requirement. It calls for a reasonably dry, hard, level, non-slippery surface, and the training materials acknowledge that standardizing the test for every type of road condition is unrealistic. Where those conditions do not exist, the original research recommended one of two things: move the test somewhere else, or administer only the eye test. Officers rarely do either.

The Four Scored Clues

There are four clues on the one-leg stand, and each has a written definition:

• Sways while balancing. A side-to-side or back-and-forth motion while you hold the one-leg position.
• Uses arms for balance. Moving your arms six or more inches away from your sides to keep your balance.
• Hopping. Keeping the raised foot up but hopping on the standing foot to stay upright.
• Puts foot down. Being unable to hold the position, and setting the raised foot down one or more times during the 30-second count.

Two of those four is a failure. Two clues, out of four, on a 30-second balance task performed on the shoulder of a road at night.

Two things about that list deserve attention. First, two of the four behaviors are never actually part of the instructions you are given. Nobody tells you not to hop, and nobody tells you not to raise your arms more than six inches. You are scored on rules you were never read. All told there are roughly nine separate things you have to do correctly while standing on one foot, and only some of them were said out loud.

Second, the list is closed. Miscounting is not a clue. Losing your place, starting over, or saying the numbers in the wrong order is not on the list, and an officer is not supposed to add to the four clues or subtract from them. If an officer testifies that you “could not even count right,” that observation has no place in the score.

What the Validation Studies Actually Found

The figure officers bring to court is a classification rate: the share of subjects an officer can correctly sort as at or above 0.08 using the two-clue threshold. That number came out of the same body of NHTSA research that produced the whole battery, and it needs three pieces of context attached before it means anything.

It is not a rate for this test alone in the field. The validation studies scored the one-leg stand as part of the full three-test battery alongside the driving and the personal contact, not as a standalone measurement.

The subjects were well past the limit. In the field validation studies, the average blood alcohol concentration of the drivers who were arrested was around 0.15, close to double the legal limit. The research is good evidence that the test identifies people who are obviously drunk. It says much less about the range where cases are actually contested. Those same studies reported a substantial share of subjects who were under 0.08 and still produced two or more clues on this test, and independent research has found that a meaningful percentage of entirely sober subjects fail the battery.

The population studied was narrow. The original research suggested that people over 65, people with back, leg, or inner ear problems, and people carrying 50 or more pounds of extra weight may have difficulty with this test. Fewer than 1.5% of the original test subjects were over 65, and the final report contained no data at all about what the subjects weighed. The research also said that anyone wearing heels more than two inches high should be offered the chance to take their shoes off. The report those findings come from is Development and Field Test of Psychophysical Tests for DWI Arrests, and it is worth reading against what an officer tells a jury the tests can do.

There is one more finding worth knowing, because it cuts the other way and is more honest for it. The original research reported that a person above 0.10 can often hold the position for as long as 25 seconds, but seldom for the full 30. The last few seconds are where this test is designed to do its work, which is also why fatigue in the standing leg is a real confound rather than an excuse.

The field validation studies conducted later in Colorado, Florida, and San Diego are where the accuracy percentages come from, and all three reported an average blood alcohol concentration around 0.15 among the drivers who were arrested.

Who Fails This Test Sober

A great many people, and for reasons that have nothing to do with drinking. Age, extra weight, an old knee or back injury, an inner ear problem, the shoes you happened to be wearing, a sloped gravel shoulder, cold, wind, headlight glare, and plain exhaustion all produce swaying and arm movement in people who have had nothing to drink at all.

Those factors belong on their own page, one at a time, with what the source actually says about each. The full list is on Who Fails Field Sobriety Tests While Sober.

The point to carry from here is narrower and it is about scoring. The four clues do not record why you swayed. They record that you swayed. A clue counted for a bad surface and a clue counted for alcohol look identical on the officer’s form.

Medical Conditions and the One-Leg Stand

The officer is supposed to ask. Before administering the balance tests, an officer is trained to find out whether you have a physical problem or disability that would affect your ability to perform, and the training materials name back, leg, and inner ear problems specifically. If you say yes, the officer has options, including alternative tasks or simply not scoring you on a test you were never a candidate for.

What happens in practice is that the question gets asked quickly, gets asked as a formality, or does not get asked at all. And when a driver does disclose a knee problem or a balance disorder, officers often proceed anyway and score the result as though the disclosure never happened.

A diagnosed condition is a different argument from an ordinary characteristic, and it is developed with medical records rather than with the video alone. That work is covered on Medical Conditions Defense.

How the Test Is Challenged in Court

The argument that works is not that the one-leg stand is a bad test in general. It is that the test given to you was not the test the research validated. NHTSA’s own materials are explicit that the accuracy claims hold only when the test is administered in the prescribed manner, the standardized clues are used, and the standardized criteria are applied. Change any element, and the validity is compromised. The officer’s own training manual supplies the standard, which is what makes this a protocol argument rather than a statistics argument.

Three administration errors show up on this test again and again, and all three make it harder than the validated version:

“If your foot comes down, pick it back up and keep counting.” This is one of the most common, and it is not part of the instructions. It is guidance to the officer about what to do if a foot comes down. Said out loud to the person being tested, it signals that putting a foot down is acceptable, when putting a foot down is one of the four clues and two clues is a failure.

“Point your toe.” Also not part of the instructions. You are supposed to be told to raise the foot and keep it parallel to the ground, which you can do with the foot hanging relaxed. Being told to point the toe engages the foot, ankle, calf, and thigh, which makes holding the position materially harder and changes the standard you are being measured against.

Bending the standing knee during the demonstration. The instructions require both legs straight. An officer who demonstrates with a slight bend in the standing leg is showing you an easier version than the one you will be graded on, because that bend acts as a shock absorber. Officers demonstrating this test have been known to put a foot down or raise their arms while doing it, which is to say they produce two clues themselves.

All of this is visible on the dashcam and bodycam footage, and none of it is visible in the offense report. How that footage gets preserved and reviewed, and where this test sits among the other two, is covered on Field Sobriety Tests.