Breath alcohol instruments: how evidential testers differ from roadside screeners
How each tester works, how models are approved and what makes a reading hold up
Safety Behind the Wheel Foundation built this guide from National Highway Traffic Safety Administration (NHTSA) guidance and product lists, federal workplace alcohol testing rules, the breath test rules of Washington, Texas and Colorado and published forensic research. Every source is listed at the end.
Why the instrument behind the number matters
A breath alcohol instrument turns a few seconds of breath into an estimate of the alcohol in a person’s body. Police agencies, investigators, road agencies and employers with safety sensitive jobs use these tools to keep impaired people from driving. Insurers and crash reconstruction specialists more often work with the results and records than run the test. In safety management, the instrument is one check among many, never the whole plan.
The number carries weight. It can lead to a license action, a criminal charge, a job decision or a finding in a car accident case. A bad reading can leave an impaired driver on the road. It can also harm a sober person. Both outcomes hurt families, so the equipment, training and records behind a result deserve a careful look.
Our guide to DUI and DWI laws by state covers limits and penalties. This article stays with the equipment: the main types, how they are approved and how a careful test is run.
Two kinds of tester, two different jobs
A preliminary tester, also called a screening device, gives a fast answer at the roadside or the workplace. An evidential breath tester, or EBT, is built to produce a result that can serve as evidence. NHTSA, the federal traffic safety agency, writes model specifications for both groups [1]. Some screening devices test saliva instead of breath [10].
Comparison compiled by Safety Behind the Wheel Foundation
Preliminary and evidential breath testers compared
| Feature | Preliminary (screening) tester | Evidential breath tester |
|---|---|---|
| Main job | Quick check to guide the next step | Result meant to be used as evidence |
| Federal workplace rule | May screen, never confirm [11] | Only type allowed to confirm [11] |
| Built in features | Maker’s quality plan, no expired devices [11] | Printed results, unique test number, air blank, external check [11] |
| Who runs it | Officer or screener on the spot | Certified operator following a set protocol [4, 5] |
A Finnish study compared 1,875 roadside screening tests with blood results from the same drivers. The screeners worked well for picking out drivers above the legal limit. But their readings scattered widely when used to estimate an exact blood level [12]. A screening result is therefore a first step, not a final answer.
Some states build that split into their rules. Washington’s state toxicologist code gives preliminary breath testing its own chapter, apart from the breath test program [3].
How a breath instrument estimates blood alcohol
Alcohol passes from blood into the air in the lungs. Deep lung breath carries a small share of the alcohol in the blood. The instrument measures alcohol in that breath, and a fixed conversion links the result to blood alcohol concentration, or BAC [13]. Some instruments use infrared sensors. Others use electrochemical sensors, which react to alcohol vapor [9].
That conversion is a legal choice, not a measurement of each person. Lawmakers set breath limits by dividing the blood limit by an assumed average ratio. Countries have used 2000:1, 2100:1, 2300:1 and 2400:1 [13]. In a lab, the ratio varies by about 2 percent, but in real people it varies by 8 to 12 percent, depending on the analyzer [13].
Breathing style matters too. Breath alcohol rises during a long exhale [13]. The federal workplace procedure tells testers to coach a steady, forceful blow of at least six seconds [10]. Texas reports results in grams of alcohol per 210 liters of breath [4].
Timing also matters. One drinking study followed ten adults with severe acid reflux. During the first 90 minutes after drinking began, breath readings tended to match or beat blood readings from a vein. After the peak, blood readings were always higher than breath readings [9].
What the NHTSA Conforming Products List is, and is not
NHTSA has written model specifications for evidential breath testers, screening devices, calibrating units and ignition interlock devices [1]. Calibrating units give a known alcohol level so users can check an instrument. NHTSA also keeps a Conforming Products List for breath testers, screening devices and calibrating units that were tested and found to conform [1].
A 2017 notice shows how the breath tester list works. It updated a 2012 list for instruments that meet specifications dated September 17, 1993. It added twelve instruments. Each entry names a model and marks it for mobile use, nonmobile use or both. NHTSA said later updates would go on its website, not in the Federal Register [2].
Here is what the list is not:
- It names models, not individual units. Each unit still needs upkeep and checks.
- It does not replace state approval, operator certification or a testing protocol [3, 4].
- It does not make any single result right. Technique, mouth alcohol and timing still matter.
Check the NHTSA page for the current list before you rely on a copy. The breath tester lists it links are dated 2012 and 2017 [1].
Who chooses, buys and runs these instruments
Public agencies and employers choose and run these tools. They buy through authorized professional instrument vendors and technical service providers, who also service the equipment. Residents do not buy or install them.
Some states put one program in charge. In Texas, the scientific director of the Department of Public Safety’s breath alcohol program must approve and certify every instrument used for evidence. The director keeps an approved list by manufacturer and model, and can pull a model, or one instrument by serial number, from use [4]. Washington’s rules name the instruments approved for measuring breath alcohol [3].
A 2005 report from New Mexico shows what approval can involve. The state laboratory, which oversees breath and blood testing for police, approved a portable instrument only after tests of accuracy, temperature, altitude, interfering substances and field use [14].
Insurers, investigators and reconstruction specialists often work from the paper trail. A careful review of a result starts with four questions. Which model was used? Who ran it? When was it last checked? What does the printout show?
How a careful evidential test is run
Steps differ by state, but good programs share a pattern. This routine blends the federal workplace procedure with rules from Texas and Colorado.
- Confirm who is being tested and explain the steps. The workplace procedure calls for photo identification and a signed form [10].
- Clear the mouth of loose material, then start the observation period. The operator stays with the person, who must not put anything in the mouth. Texas requires at least 15 minutes. Colorado requires 20 and starts over if the person belches, regurgitates or takes anything into the mouth [4, 5].
- Use a fresh mouthpiece for every sample [5]. The workplace procedure calls for a sealed one [10].
- Coach a steady, forceful blow of at least six seconds under the workplace procedure [10].
- Take two samples. Colorado leaves two minutes between them and keeps the person away from the instrument [5]. Duplicates help expose mouth alcohol [6].
- Check the reference system, a sample of known strength run alongside the test. In Texas it must agree with the known value within 0.01 gram per 210 liters, or limits the director sets [4].
- Keep the printout and records. Workplace breath testers print results with a unique test number and the test time [11].
In workplace testing, a screening result of 0.02 or more leads to a confirmation test on an evidential tester. The person is told not to eat, drink, put anything in the mouth or belch while waiting [10, 11].
What can go wrong with a reading
Several common problems come from the setting and the sample, not from the machine itself. Trained operators plan for these.
Comparison compiled by Safety Behind the Wheel Foundation
Common problems with breath readings
| Problem | What can happen | How programs respond |
|---|---|---|
| Mouth alcohol | Reading runs high | Wait 15 to 20 minutes and take two samples [4, 5, 6] |
| Mouthwash or sanitizer vapor | False positive or a status code | Observation period, no sanitizer beforehand [7, 8] |
| Short or uneven blow | Reading changes with blow length | Coach a full, steady blow [10, 13] |
| Screening used as proof | Wide scatter in exact level | Confirm on an evidential tester [11, 12] |
Mouth alcohol is the classic problem. One study gave seven people who had been drinking a rinse with a vodka solution. Their readings fell by more than 0.020 gram per 210 liters between the first and second tests. The reading took about nine minutes on average to settle, and no more than 13 [6]. In an 11 person study, breath alcohol after an ethanol mouthwash took 6 to 15 minutes to fall below Sweden’s limit [8].
Hand sanitizer on an operator’s hands can matter too. In one study, 13 of 130 first analyses (10 percent) read positive, and 41 (31.5 percent) triggered a status code, a warning that the instrument flagged the sample [7]. Repeat samples never gave back to back positive numbers. The authors advised operators to skip sanitizer for 15 minutes before testing [7]. Safety Behind the Wheel Foundation gives testing programs the same advice.
After a car accident, a test may happen hours later. Experts sometimes calculate backward to estimate an earlier level. That method assumes alcohol absorption finished within about two hours of the last drink. Food, some medicines and medical conditions can stretch absorption longer, so toxicologists should say that population averages may not fit one person [15].
Upkeep, records and the questions to ask
Like any measuring tool, a breath instrument needs regular checks against a known standard. Under the workplace procedure, users follow the maker’s instructions, run external checks with calibrating units from the NHTSA list, and keep inspection and maintenance records. Screening devices follow the maker’s quality plan and cannot be used past their expiration or after failing a quality check [11].
Labs check instruments with a simulator, which gives off alcohol vapor at a known strength, or with tanks of reference gas [14]. Texas lets its scientific director inspect testing sites and requires records [4].
Federal and state rules in brief
As of October 2026, federal workplace rules for transportation employers fix the equipment. Only devices on the U.S. Department of Transportation’s lists may screen, and only evidential breath testers may confirm [11]. Beyond that setting, each state sets its own program for evidential testing in impaired driving cases.
Three states show how much the details differ.
Comparison compiled by Safety Behind the Wheel Foundation
Breath test rules in three states
| State | Who sets the rules | Example detail |
|---|---|---|
| Washington | State toxicologist | Own chapter for preliminary testing, and a rule names approved instruments [3] |
| Texas | Public safety department’s scientific director | Approves instruments, 15 minute wait, results in grams per 210 liters [4] |
| Colorado | Health department’s laboratory division | Separate rules for operators, sites and procedures, plus a 20 minute wait [5] |
Even basics vary. The wait is 15 minutes in Texas and 20 in Colorado [4, 5]. In one small study, mouth alcohol cleared within 13 minutes in every subject [6]. So both fall on the careful side. Your state may differ again, so check the source for your own state.
What you can do as a resident, employer or local official
Public traffic safety equipment belongs in the hands of authorized agencies. Here is where ordinary people fit.
- Drivers and families: A breath tester bought for personal use is a personal check. Unless it is on your state’s approved list and run under the state’s procedure, it is not an evidential instrument. A low reading does not prove you are safe to drive, so plan a sober ride. If an officer asks you to take a test, ask which kind it is, and learn your state’s rules before you drive. Our guide to alcohol impaired driving shows how.
- Employers: If your drivers fall under federal transportation rules, use listed devices and follow the federal procedure. Otherwise state law decides, so ask a qualified service provider and keep records. See our guide to workplace driving impairment policies.
- Residents and officials: Ask your police agency or state breath test program how instruments are approved, who certifies operators and how often instruments are checked. Use the questions in the checklist above.
- Anyone who has been tested: Ask the agency for the proper way to request records. For legal questions, talk to a licensed attorney in your state.
Frequently asked questions
Is the roadside breath test the same as the test at the station?
Usually not. A roadside device is a preliminary tester that helps an officer decide the next step. The station test uses an evidential instrument, run by a certified operator under a set protocol. Federal workplace rules follow the same split: screening devices may screen but never confirm [11]. Washington even regulates the two in separate chapters [3].
Does being on the NHTSA list mean a result is correct?
No. The list says a model was tested and found to conform to NHTSA’s model specifications [1]. A result still depends on the unit’s upkeep, the operator’s technique and the person’s condition at the time. States add their own approvals and procedures [3, 4].
Can mouthwash or hand sanitizer change a breath test?
Yes, if the test comes too soon. Ethanol mouthwash gave a false positive when breath was sampled within 15 minutes [8]. Sanitizer vapors caused positive readings or status codes in some first samples [7]. That is why programs use a waiting period and a second sample.
Can acid reflux cause a false high reading?
A small study of ten adults with severe reflux found that this risk was highly improbable. Four of the ten had reflux during the study, yet readings did not stray widely from blood results [9]. One small study cannot settle every case, so ask the program how it handles medical conditions.
Why does a breath test need two samples and a wait?
Both guard against mouth alcohol. In one study, readings fell sharply between the first and second tests after a rinse [6]. Texas requires 15 minutes of continuous presence, and Colorado requires 20 [4, 5].
Can a test taken hours after a crash show the level at the time of driving?
Only as an estimate. Backward calculations assume a standard absorption time and rely on averages. Food, some medicines and medical conditions can change absorption, so experts should state those limits [15].
The bottom line
Recommendation from Safety Behind the Wheel Foundation
Safety Behind the Wheel Foundation urges everyone to look past the number to the instrument, the operator and the records, and never to drive after drinking. A breath number is only as sound as the instrument, the operator and the process behind it. Screening devices point the way. Evidential instruments, approved and checked under state and federal rules, produce the result others rely on.
You do not need to run these tools to ask good questions. Ask which type was used, whether it is approved, who ran it and where the records are. Better still, never drive after drinking. No instrument makes a drinking driver safe.
Related articles
Sources and further reading
- 1National Highway Traffic Safety Administration (NHTSA). Alcohol Measurement Devices. Page reviewed October 2026; no update date shown. nhtsa.gov/…/alcohol-measurement-devices
- 2National Highway Traffic Safety Administration (NHTSA). Highway Safety Programs; Conforming Products List of Evidential Breath Alcohol Measurement Devices (Federal Register notice). Published November 2, 2017. nhtsa.gov/node/37041
- 3Washington State Toxicologist (Washington Administrative Code, as published by Cornell Legal Information Institute). Title 448 WAC, State Toxicologist and “WAC 448-16-020, Approval of breath test equipment.” Current text reviewed October 2026. [https://www.law.cornell.edu/regulations/washington/title-448](https://www.law.cornell.edu/regulations/washington/title-448) and law.cornell.edu/…/WAC-448-16-020
- 4Texas Department of Public Safety (Texas Administrative Code, as published by Cornell Legal Information Institute). 37 Tex. Admin. Code section 19.2, Instrument Certification and “section 19.3, Techniques and Methods.” Current text reviewed October 2026. [https://www.law.cornell.edu/regulations/texas/37-Tex-Admin-Code-SS-19-2](https://www.law.cornell.edu/regulations/texas/37-Tex-Admin-Code-SS-19-2) and law.cornell.edu/…/37-Tex-Admin-Code-SS-19-3
- 5Colorado Department of Public Health and Environment, Laboratory Services Division (Code of Colorado Regulations, as published by Cornell Legal Information Institute). 5 CCR 1005-2, Testing for Alcohol and Other Drugs, including Part 4, sections 4.2 and 4.3 (Evidential Breath Alcohol Testing procedures). Current text reviewed October 2026. [https://www.law.cornell.edu/regulations/colorado/5-CCR-1005-2-4.2](https://www.law.cornell.edu/regulations/colorado/5-CCR-1005-2-4.2) and [https://www.law.cornell.edu/regulations/colorado/5-CCR-1005-2-4.3](https://www.law.cornell.edu/regulations/colorado/5-CCR-1005-2-4.3) (rule contents: [https://www.law.cornell.edu/regulations/colorado/department-1000/division-1005/5-CCR-1005-2](https://www.law.cornell.edu/regulations/colorado/department-1000/division-1005/5-CCR-1005-2)) law.cornell.edu/…/5-CCR-1005-2
- 6Sterling K. The rate of dissipation of mouth alcohol in alcohol positive subjects. Journal of Forensic Sciences 57(3):802-805. Published online December 28, 2011. pubmed.ncbi.nlm.nih.gov/22211670
- 7Strawsine E, Lutmer B. The Effect of Alcohol-Based Hand Sanitizer Vapors on Evidential Breath Alcohol Test Results. Journal of Forensic Sciences 63(4):1284-1290. Published online November 16, 2017. pubmed.ncbi.nlm.nih.gov/29143333
- 8Ernstgard L, Pexaras A, Johanson G. Washout kinetics of ethanol from the airways following inhalation of ethanol vapors and use of mouthwash. Clinical Toxicology 58(3):171-177. Published online June 19, 2019. pubmed.ncbi.nlm.nih.gov/31215247
- 9Kechagias S, Jonsson KA, Franzen T, et al. Reliability of breath-alcohol analysis in individuals with gastroesophageal reflux disease. Journal of Forensic Sciences 44(4):814-818. Published July 1999. pubmed.ncbi.nlm.nih.gov/10432616
- 10U.S. Government, Electronic Code of Federal Regulations. 49 CFR Part 40, Subpart L: Alcohol Screening Tests (sections 40.241-40.247). Current text reviewed October 2026. ecfr.gov/…/subpart-L
- 11U.S. Government, Electronic Code of Federal Regulations. 49 CFR Part 40, Subpart K: Testing Sites, Forms, Equipment and Supplies Used in Alcohol Testing (sections 40.221-40.235). Current text reviewed October 2026. ecfr.gov/…/subpart-K
- 12Kriikku P, Wilhelm L, Jenckel S, et al. Comparison of breath-alcohol screening test results with venous blood alcohol concentration in suspected drunken drivers. Forensic Science International 239:57-61. Published 2014. Data year: three-month sampling period in Finland (year not stated in abstract). pubmed.ncbi.nlm.nih.gov/24747668
- 13Jones AW. Breath-alcohol analysis as a surrogate for blood-alcohol concentration depends on assuming a constant blood/breath ratio of alcohol. Forensic Science Review 37(2):117-135. Published July 2025. pubmed.ncbi.nlm.nih.gov/40750099
- 14Razatos G, Luthi R, Kerrigan S. Evaluation of a portable evidential breath alcohol analyzer. Forensic Science International 153(1):17-21. Published October 2005. pubmed.ncbi.nlm.nih.gov/15935582
- 15Olson A. Extended absorption, implications: Rethinking alcohol pharmacokinetics in forensic calculations. Medico-Legal Journal 94(1):24-27. Published online November 18, 2025. pubmed.ncbi.nlm.nih.gov/41254916
This article reflects the safety priorities of impaired driving prevention and accurate, fair alcohol testing. It is general education, not legal or medical advice. Rules change and differ by state, so verify details with official sources. For your own situation, contact your state's breath test or toxicology program, your state DMV, your employer's testing contact or a licensed attorney in your state. Reviewed October 2026.
Cite this article: Safety Behind the Wheel Foundation. “Breath alcohol instruments: how evidential testers differ from roadside screeners.” drivewithcare.org, https://drivewithcare.org/evidential-vs-preliminary-breath-alcohol-testers/. Reviewed October 2026.
Special thanks to our contributors
We appreciate the professionals, experts, and community members who share their knowledge and perspectives to help make this resource better. Your insights are always welcome and appreciated.
Have professional expertise, firsthand experience, or a perspective that could make this resource better?
We welcome your input. Your knowledge and experience can help us make this resource more useful, accurate, and meaningful to the people who rely on it.
Thank you. Your contribution is on its way.
We’re grateful you took the time to share your knowledge with us. Our team reviews every contribution with care. If it helps improve this resource, we’ll update the page and recognize you the way you chose.
Keep learning
Safe Driving Resource Center
Guides, self-checks and articles for drivers, parents of new drivers, and families recovering after a crash.
