The walk and turn is a divided attention exercise. It asks you to do two things at once, balance in an awkward stance while keeping a set of instructions in your head, on the theory that alcohol degrades the ability to split attention. Under State v. Meador, 674 So. 2d 826 (Fla. 4th DCA 1996), the officer can describe how you performed as a lay observation, but cannot tell the jury you scored points or failed, because that borrows a scientific certainty the exercise does not have.
There are eight clues, and the often-missed detail is that they are split across two stages that the officer grades separately. The instruction stage begins the moment you are told to stand heel to toe with your arms at your sides and listen. You are already being scored before you take a single step, on whether you can hold that position and whether you start too soon. The walking stage is the nine heel-to-toe steps out, the turn, and the nine steps back.
Eight clues across two separate stages. NHTSA scores the standing instruction stage and the walking stage independently, which many folks do not realize while they are being graded on both.
Two or more clues is the whole case, and the surface decides it
The training treats two or more clues out of eight as the decision point for likely impairment. That is a low bar, and it is why the conditions matter so much. NHTSA requires a designated straight line, real or clearly drawn, on a reasonably dry, hard, level, and non-slip surface. The roadside shoulder, with its slope, gravel, paint, and seams, is rarely that. Stepping off an imaginary line, using your arms on a slope, or missing heel to toe in stiff shoes are clues that say more about where you were standing than about alcohol.
Two of the eight clues happen before you take a step
The part that surprises many folks is where the scoring starts. Two of the eight clues, losing your balance while the officer talks and starting before you are told to, are scored during the instructions, while you are standing heel to toe on a roadside listening to a stranger recite a procedure you have never done. Two clues is the decision point for the whole exercise, so a person can reach it without ever taking a step, on nerves and an awkward stance alone.
| Clue | Stage | A sober reason it happens |
|---|---|---|
| Cannot keep balance while listening | Instructions | An unnatural heel-to-toe stance held on uneven ground |
| Starts too soon | Instructions | Nerves, or instructions that ran together |
| Stops while walking | Walking | Steadying yourself on a dark shoulder |
| Does not touch heel to toe | Walking | Footwear, age, or a knee or back problem |
| Steps off the line | Walking | There was no real line to stay on |
| Raises arms for balance | Walking | A normal reflex when balancing on one line |
| Turns improperly | Walking | A turn few people have ever practiced |
| Wrong number of steps | Walking | Miscounting under stress, not intoxication |
Two of these end the exercise. Set next to the ordinary reasons each one happens, that is a thin line to build a case on.
Who should never be graded on this exercise
The procedure itself carves out people for whom the exercise was never meant. The guidance is that people who are over about 60, who are more than 50 pounds overweight, or who have back, leg, or inner-ear conditions that affect balance should not be given the walk and turn, and that anyone in heels over two inches should be allowed to take their shoes off. When an officer runs a 64-year-old with a bad knee through it on a sloped shoulder and then counts the arm raise as a clue, the clue is an artifact of the setup.
Demonstrations and instructions are part of the test
The officer is required to demonstrate the stance and the steps and to give the instructions in a specific way. A demonstration that is skipped or done wrong, or instructions that are rushed or garbled, are not minor. They change what you were asked to do, which means the clues that follow were scored against a standard you were never properly given. As an NHTSA-qualified instructor, I compare the bodycam to the exact script and walk a jury through every place the administration drifted.
On the walk and turn I start where the officer started, with the instructions, and I do it with you, not at you. I show the jury the two clues that get counted before you take a single step, and I go clue by clue against the video, your shoes, your age, and your own body. I teach this exercise, so I can hold the officer to the exact procedure they were trained to give. When the line was imaginary and the surface was wrong, a two-clue call comes down to a handful of ordinary movements written against a nervous person on the side of a road, and I make sure the jury sees exactly that.
I started out as an Assistant Public Defender in Florida’s Thirteenth Judicial Circuit, in Tampa, and today I am one of six ACS-CHAL Forensic Lawyer-Scientists in Florida and an NHTSA-qualified field sobriety instructor. I was trained to administer this battery and to teach the officers who give it, so I can stand in front of a jury and show the exact point where the roadside exercises in your case left the standard. Learn more about my background.
The 68 Percent Problem
The scoring rule for this exercise comes from NHTSA’s validation research: two or more of the eight clues is treated as a failed test. That decision rule classified about 68 percent of people correctly in the original research against a 0.10 standard, and the 1998 San Diego study published 79 percent at 0.08. Roughly one person in three at the original threshold was sorted wrong by the very standard the officer applies at the roadside.
It gets worse for drivers who were under the limit. The San Diego study never published its false positive rates, but they can be calculated from its own tables: of 76 tested drivers below 0.08, the walk and turn flagged 40 as over. That is a 52.6 percent false positive rate. A coin flip would have treated sober drivers more fairly, and that comes from the government’s flagship validation study, not from a defense expert.
Two clues is a low bar. Starting before the instructions finish is a clue. Raising your arms for balance is a clue. On that math, a nervous, sober driver on a sloped shoulder at 2 a.m. can hit the arrest threshold without a drop of alcohol, which is why the bodycam of the instruction phase gets as much of my attention as the walking itself.
Questions About the Walk and Turn
What is the walk and turn test?
A divided attention exercise. You stand heel to toe and listen to instructions, then take nine heel-to-toe steps out, turn, and take nine back. The officer watches for eight clues across the standing and walking parts, on the theory that alcohol hurts the ability to do two things at once.
How many clues are there on the walk and turn?
Eight, and they are split between two stages graded separately. Two clues come from the instruction stage and the rest from the walking stage. The training treats two or more total as the decision point, which is a low bar.
Does the surface matter?
A great deal. NHTSA requires a designated straight line on a reasonably dry, hard, level, non-slip surface. A sloped, graveled, or painted roadside shoulder produces stepping off, arm raises, and missed heel-to-toe that say more about the ground than about alcohol.
Should everyone be asked to do it?
No. The procedure says people over about 60, more than 50 pounds overweight, or with back, leg, or inner-ear problems should not be given it, and people in heels over two inches should be allowed to remove their shoes. Grading someone the exercise excluded is a strong challenge.
What if the officer skipped the demonstration?
That matters. The officer must demonstrate the stance and steps and give the instructions a specific way. A skipped or wrong demonstration changes what you were asked to do, so the clues were scored against a standard you were never properly given.
How do you challenge it in my case?
I compare the bodycam to the exact NHTSA script, check the surface, your footwear, and any physical condition, and show the jury each place the administration drifted from the standard the exercise was validated under.
Related: the one leg stand, the HGN eye test, administration errors, the surface and the roadside, and how clues are scored.
Can I fail before I even start walking?
In effect, yes. Two of the eight clues are scored during the instructions, when you are holding a heel-to-toe stance and listening, and two clues is the decision point, so nerves and an awkward stance alone can reach it before you take a step.
This page is general information, not legal advice, and it does not create an attorney-client relationship. Field sobriety exercises in Florida are governed by case law and by section 316.193, Florida Statutes. Procedures and rules change, and every case turns on its own facts. Past results do not guarantee a similar outcome.
