Ask Deandra: Do I Have to Take a Field Sobriety Test in Texas?

Here is the longer answer: what field sobriety tests actually are, why they are voluntary, why their reliability is overstated, and what happens if you decline.

What Field Sobriety Tests Are

“Field sobriety tests” means the three Standardized Field Sobriety Tests, or SFSTs, developed by the National Highway Traffic Safety Administration in the 1970s and 1980s and still used today. The three tests are:

• Horizontal Gaze Nystagmus (HGN). The officer asks you to follow a stimulus, a pen or a finger, with your eyes only, and watches for involuntary jerking of the eye.
• Walk-and-Turn (WAT). The officer asks you to take nine heel-to-toe steps along a real or imaginary line, turn in a specific way, and take nine more steps back, while counting aloud and following specific instructions.
• One-Leg Stand (OLS). The officer asks you to stand on one leg with the other foot held six inches off the ground, count out loud, and hold the position for thirty seconds.

Officers may also ask you to perform tests that are not part of that battery, such as reciting the alphabet, counting backwards, or touching your finger to your nose. None of these were validated. NHTSA says so about the first two in its own training material, which describes them as optional techniques that have not been scientifically validated. Finger-to-nose is a more interesting case: it was one of the six candidate tests in the original research and it was dropped, because the studies found the three that survived were the accurate ones and the rest were measuring the same skills over again. The Romberg balance test that officers sometimes use is not in the SFST program at all; it comes out of drug recognition training. All of them are voluntary, exactly like the three standardized tests.

Each of the three standardized tests has its own clue list, its own scoring threshold, and its own set of conditions that produce failure in a sober person. We cover each one separately: the HGN eye test, the walk-and-turn test, and the one-leg stand test.

Why They Are Voluntary

Texas law does not require a driver to perform field sobriety tests. There is no statute analogous to the implied consent law that governs breath and blood testing. There is no separate offense for refusing. There is no automatic license suspension. The tests are presented to you in a way that suggests refusal is a problem (“I just need you to do a few things for me”) but you have the legal right to say no.

This is one of the most important pieces of information in DWI defense, and one of the most consistently underexplained to the public. Officers are not required to tell you the tests are voluntary, and most do not. The presumption built into the encounter is that you will perform them. The legal reality is that you do not have to. If you want the whole roadside script, we set it out in what to do when you are pulled over.

Why the Officer Is Asking

The roadside exercises sit at a specific point in the encounter, and understanding that point explains why they exist at all.

By the time an officer asks you to step out of the car, he already has reasonable suspicion. That is what justified the stop and the continued detention. What he does not yet have is probable cause to arrest, which is a higher standard. The field sobriety tests are the bridge between the two. They are the mechanism by which an officer converts a suspicion into an arrest decision, and they are designed to generate a scored, video-recorded record that supports that decision after the fact.

That is why the request comes when it does, and why it is framed as a small favor rather than as evidence collection. The officer is not gathering information to decide whether to let you go. He is building the record for a decision that is usually already leaning one direction.

How the Three Tests Are Scored

The tests are not pass or fail in the way most people picture. Each one has a fixed list of specific behaviors, called clues, and a numerical threshold at which NHTSA says the officer may conclude the subject is at or above 0.08.

• Horizontal Gaze Nystagmus. Six clues, three observed in each eye: lack of smooth pursuit, distinct and sustained nystagmus at maximum deviation, and onset of nystagmus prior to 45 degrees. The decision point is four clues.
• Walk-and-Turn. Eight clues: cannot keep balance during the instructions, starts too soon, stops while walking, does not touch heel to toe, steps off the line, uses arms for balance, improper turn, and wrong number of steps. The decision point is two clues.
• One-Leg Stand. Four clues: sways while balancing, uses arms for balance, hops, and puts the foot down. The decision point is two clues.

Read those thresholds again, and read the definitions with them. On the walk-and-turn, “uses arms for balance” means raising either arm six inches from your side, and “does not touch heel to toe” means a gap of half an inch. Two of those on a sloped gravel shoulder is a scored failure. On the one-leg stand, swaying and putting your foot down once in thirty seconds is enough. The margin between a clean performance and a scored failure is much narrower than the phrase “failed the field sobriety tests” suggests to a jury.

It is worth knowing that NHTSA agrees with that framing. Its own training material tells officers that the SFSTs are a tool to help them see visible signs of impairment and are not a pass or fail test. The officer who testifies that you failed is using language his own manual tells him not to use.

The thresholds also assume the test was administered correctly. NHTSA specifies the conditions: for the two walking and balancing tests, it recommends a reasonably dry, hard, level, non-slippery surface, adequate lighting, and standardized verbal instructions delivered in a prescribed sequence with a demonstration. Its guidance is that if any one of the SFST elements is changed, the validity may be compromised. Whether those conditions were met is a factual question answered by the video, not by the offense report.

The clue counts, the clue definitions, and the decision points are not our characterization of the tests. They come from the standardized battery NHTSA developed in Development and Field Test of Psychophysical Tests for DWI Arrests and carried forward into the training manual the officer was certified on. That manual is the document the administration gets measured against, and obtaining the edition in force on the night of the arrest is part of the work.

The Eye Test Has Its Own Rules

Of the three tests, the eye test is the one Texas law treats differently, and the differences matter to anyone whose case includes one.

The walk-and-turn and the one-leg stand come into evidence as ordinary observation. The officer describes what he saw about your coordination and balance, and Texas courts treat that as the kind of thing any witness could describe. HGN is not like that. Texas courts classify it as scientific evidence, which means it has to satisfy the reliability standard the courts apply to science. The Court of Criminal Appeals decided in 1994 that the theory behind HGN and the technique are reliable enough to be taken as given, so the State does not have to prove those from scratch in every case.

What the State does still have to prove is that the test was properly applied on your night, and that the officer was qualified to give it. On qualification the rule is specific: the witness has to be qualified as an expert on the HGN test, and for a peace officer that is satisfied by proof of Texas practitioner certification. This is why the certification records matter and why a lapsed or missing certification is worth chasing down.

Two more points, and the first is the one most often gotten wrong. An officer is not allowed to turn your eye movements into a number. The Court of Criminal Appeals held that a qualified witness may describe your performance on the HGN test but may not correlate that performance to a particular blood alcohol concentration, because NHTSA’s own formula for estimating BAC from the angle of onset carried a margin of error the court found too large to allow. If an officer testifies that your eyes told him you were at a specific level, that is not permitted.

Second, the vertical version of the eye test is a different animal. Vertical gaze nystagmus is taught alongside HGN and administered with it, but it is not one of the three validated tests, it has no clue count and no decision point, and Texas courts have held that the 1994 decision approving HGN does not carry over to it. The science behind vertical nystagmus, and behind resting nystagmus, has to be established separately, and without that showing the evidence does not come in. That is a distinction worth raising at a hearing on the reliability of the evidence.

Why the Tests Are Less Reliable Than They Look

Field sobriety tests are presented as objective scientific assessments. They are not. They are observational tests scored by the officer based on the officer’s perception of “clues”. Clues are specific behaviors NHTSA has linked, with varying degrees of validation, to alcohol impairment. Each test has well-documented limitations:

• HGN problems. Nystagmus has many causes other than alcohol, including fatigue, certain medications, neurological and eye conditions, head injuries, exposure to certain chemicals, even high caffeine intake. The administration of the test, meaning stimulus distance, angle, speed of movement and head positioning, has to follow precise protocols, and deviations are common. Body camera and dashcam review frequently shows administration that does not meet the NHTSA standard.
• Walk-and-Turn problems. NHTSA recommends a flat, dry, hard, non-slippery surface with enough room for nine heel-to-toe steps. The reality of a roadside stop is rarely that. NHTSA’s own manual flags people over 65 and people with back, leg or inner ear problems as likely to have difficulty with this test, and advises that anyone in heels more than two inches high be given the chance to take their shoes off. Fatigue, injury and unfamiliar footwear do the rest.
• One-Leg Stand problems. The same age and physical-condition caveats apply, and this test carries one more: NHTSA’s manual flags people who are 50 or more pounds overweight as likely to have difficulty. That caveat belongs to the one-leg stand specifically, and it is worth being precise about, because the manual also concedes that the original studies did not record subject weight. Anyone who has tried to balance on one foot in a stressful environment knows that thirty seconds is longer than it sounds.
• Roadside conditions. Flashing emergency lights, traffic noise, weather, time of night, and the sheer stress of being investigated for DWI all affect performance. Be aware that NHTSA has taken a position here: its manual says field validation studies indicated varying environmental conditions did not affect subjects’ ability to perform, and a prosecutor will have that sentence ready. The answer is not that the conditions void the test. It is that the conditions are one of the things the officer was supposed to account for and usually did not, and the video shows whether he did.

None of this means SFSTs are useless. It means they are observational evidence subject to interpretation, not chemical evidence with a defined error rate. The State will present a defendant’s SFST performance as objective proof of intoxication. A forensic-trained defense lawyer will present the same performance as an unreliable observational test administered under poor conditions to a person whose innocent factors were never accounted for. The specific conditions that cause sober people to produce clues, including age, weight, footwear, surface, lighting and medical history, are covered in conditions that cause sober people to fail field sobriety tests.

What the Validation Research Actually Claims

The accuracy figures the State relies on come from a small set of NHTSA-sponsored field validation studies, and they are narrower than the way they get presented in a courtroom.

The most frequently cited is the 1998 San Diego field validation study, which reported officer accuracy in identifying subjects at or above 0.08 at about 88 percent for HGN, 79 percent for walk-and-turn and 83 percent for one-leg stand, with all three used together at about 91 percent. The two earlier studies, in Colorado and Florida, reported in a similar range, though they measured slightly different things: Colorado reported 86 percent accurate arrest and release decisions, and Florida reported correct decisions to arrest 95 percent of the time. These are the most quoted numbers in this area and also the most frequently misstated in a courtroom, which is why the underlying reports are worth having rather than the summary an officer recites.

Two things about those numbers matter more than the numbers themselves.

First, and this is the study’s own point rather than ours, the tests were designed to indicate a driver’s blood alcohol concentration rather than to indicate driving impairment, and the San Diego report says in terms that the link between the two is a separate question requiring entirely different research. That distinction is not a technicality in Texas. A Texas jury is asked whether the defendant had lost the normal use of mental or physical faculties. The validation research was never aimed at that question.

Second, the figures describe whether an officer correctly sorted a subject onto one side of a line, under study conditions, with administration that was monitored. That is not the condition of a routine roadside stop at one in the morning on a sloped shoulder with traffic passing. It is also worth noting what the sample was: drivers the officers had already decided to stop. The number is an accuracy rate within an existing suspicion, not an error rate for the population of drivers on the road.

How those studies were designed, who funded them, what the methodological criticisms are, and how the tests were adopted nationally anyway is covered in our critical history of the SFSTs.

What Happens If You Refuse

If you decline to perform field sobriety tests at the roadside, here is what does and does not happen:

• You are not charged with refusing. There is no separate offense for declining SFSTs in Texas.
• Your license is not automatically suspended. The Administrative License Revocation process is triggered by refusal of a breath or blood test, not by refusal of field sobriety tests. That is a genuine difference between the two decisions and it is why they should not be thought about the same way.
• The officer may still arrest you. If the officer believes there is probable cause for DWI based on driving behavior, observations, statements and other indicators, the officer can arrest you without SFSTs.
• Your refusal can be used at trial. This is the part to understand before you decide. Texas courts have held that evidence of a refusal to perform field sobriety tests is admissible, reasoning from the same principle that lets the State prove a breath test refusal. Declining is not taking the Fifth, because the courts treat these tests as physical rather than testimonial. So the prosecutor is allowed to tell the jury you refused and to argue about what that means. The defense answer is straightforward: the tests are voluntary, the defendant exercised a legal right, and declining an unreliable observational test is not an admission of intoxication.

The trade-off is real but generally favorable to the defendant. By declining, you remove a category of evidence the State would otherwise use, in exchange for an argument the prosecution gets to make about the refusal. In most cases that is a trade worth making, because once you perform the tests, the officer’s scoring of your performance, on video, becomes part of the State’s case against you, and a video of someone struggling is far more persuasive to a jury than a sentence about a refusal.

How to Decline

Politely and consistently. The same script that works for declining other roadside requests works here:

• “Officer, I respectfully decline to perform any tests.”
• “I’d prefer not to do that without a lawyer.”
• “I understand you’re asking, but I’d rather not.”

Do not argue. Do not lecture the officer about the validity of NHTSA studies. Do not explain why you would have failed even sober. Decline once, decline again if asked again, and stop talking. Say the words rather than simply going quiet, because silence you never label as an invocation is just silence. The video should show a calm, polite person exercising their rights, not a person trying to talk their way out of a stop.

What About the Portable Breath Test (PBT)?

After or during the SFSTs, officers often ask the driver to blow into a small handheld device on the side of the road. This is the portable breath test, or PBT. It is a screening device, not the evidentiary instrument used at the station. The roadside PBT is voluntary in Texas, and you can decline it the same way you decline SFSTs.

It is also worth knowing that the number a roadside device produces will never be shown to a jury. Texas courts hold that portable breath test results are not reliable enough to be admitted, though an officer may still use the reading as one factor supporting probable cause to arrest. Like the field sobriety tests, it is a tool that helps the officer decide, not evidence that helps the State prove.

Declining the PBT is a separate decision from declining the breath or blood test at the station. The DIC-24 statutory warning, the implied-consent framework, and the ALR consequences all apply to the station test but not the roadside PBT.

If You Have Already Performed Field Sobriety Tests

If you took the tests before reading this post or talking to a lawyer, you are not without recourse. SFST evidence is one of the most challengeable categories of evidence in a DWI case. Defense counsel can:

• Review the video. Body cam and dashcam footage usually tells a different story than the officer’s narrative report. Administration errors, scoring inconsistencies, and conditions that confound the test are routinely visible.
• Examine the officer’s training and certification records. SFST certification has expiration dates, refresher requirements, and documented standards. Lapses and gaps are common, and on the eye test in particular the certification is not a formality: it is what makes the testimony admissible in the first place.
• Develop the medical and physical issues that affected performance. Age, weight, back and knee problems, vestibular issues, and neurological or eye conditions can all be developed through medical records and expert testimony.
• Cross-examine on the conditions of administration. Surface, lighting, weather, traffic, footwear, and the officer’s adherence to NHTSA protocol are all fair game. How far a deviation has to go before it matters is itself a live question: Texas courts have treated slight variations as going to the weight of the evidence, while a record that does not establish compliance at all has supported keeping the test out.
• Look at whether the stop itself was lawful. Everything downstream depends on it. If the stop was bad, what the tests showed may not matter.

Performing the tests is not the end of the case. It just means the defense work has to be more thorough. The full picture of how SFST evidence gets attacked, from the certification records through the video review to the cross-examination itself, is on challenging field sobriety test results.

The Bottom Line

Field sobriety tests are voluntary in Texas. They are observational, subjective, and far less reliable than the State will lead a jury to believe. Declining them removes a category of evidence the prosecution will otherwise use, while exposing the defendant to only a manageable trial argument about why the refusal occurred. If you can remember anything before a DWI stop ever happens, remember this: identify yourself, hand over your license, decline the tests, ask for a lawyer, and stop talking. The roadside is the wrong place to try to win a DWI case. The right place is a courtroom with a defense lawyer who knows the science.

For a more detailed history of the development of the SFSTs, read our guide to the history and science of field sobriety tests.

DWI Defense at Deandra Grant Law

Deandra Grant Law defends DWI and intoxication-offense cases across North and Central Texas including Dallas, Fort Worth, Plano, McKinney, Frisco, Allen, Lewisville, Denton, Rockwall, and Waco. We review SFST administration on every traffic stop video, identify the protocol failures and confounding factors that the offense report leaves out, and challenge the evidence at the level the science requires. Deandra Grant herself is an SFST instructor.

If you have been arrested for DWI in Texas, whether you took the field sobriety tests or declined them, call Deandra Grant Law at (214) 225-7117 to schedule a confidential consultation. And remember that the 15-day ALR deadline runs from the date of service of the notice of suspension.

Have a DWI question you want answered in this series? Submit it at texasdwisite.com. You might see it featured in a future Ask Deandra post.

Quick Answers: Field Sobriety Test Refusal in Texas

Are field sobriety tests mandatory in Texas?

No. The roadside exercises are voluntary, and there is no license suspension or criminal penalty for declining them, unlike a breath or blood test refusal.

Can you refuse a field sobriety test in Texas?

Yes, politely and clearly. Officers do not have to tell you the tests are optional, and declining them denies the State scored evidence it would otherwise use.

What happens if I refuse the field sobriety tests?

You may still be arrested if the officer believes there is probable cause from other observations, but the file will not contain a scored test performance, which is often the State’s most persuasive exhibit. The prosecutor is allowed to tell the jury that you declined, so the choice is a trade rather than a free one.

Can an officer testify that my eyes showed a specific blood alcohol level?

No. A certified officer may describe what he observed during the eye test, but Texas law does not allow him to convert that observation into a blood alcohol number.