To use a breathalyzer accurately, wait at least 15 to 20 minutes after your last drink (or anything else in your mouth), attach a clean mouthpiece to a warmed-up device, then exhale steadily until it signals it has enough air. The number you get is only as trustworthy as the sensor inside the device and the technique you used to produce the sample, so the details below matter more than they might seem.
Start With the Right Kind of Device
Breathalyzers use one of two sensors, and the gap between them is wide enough to change whether you can trust a reading near a legal threshold.
Semiconductor sensors measure how electrical resistance changes when breath passes over a tin-oxide chip. They’re cheap and small, and fine for confirming that yes, you’ve been drinking. They also react to compounds other than ethanol, including acetone from low-carb diets or diabetes, which makes them prone to false positives and faster calibration drift.
Fuel cell sensors oxidize the alcohol in your breath and generate a small electrical current proportional to the concentration. Law enforcement and workplace testing programs use fuel cell devices almost exclusively because they respond much more specifically to ethanol. A police-grade fuel cell unit like the LifeLoc FC10 reports accuracy within ±0.005 BAC up to a reading of 0.100.1National Center for Biotechnology Information. The Accuracy and Promise of Personal Breathalysers for Research Consumer fuel cell devices run roughly $50 to $150 and are less precise than professional models, but they still outperform semiconductor sensors by a wide margin.
If you’re going to make decisions based on the number, buy a fuel cell model. Everything below assumes you have one.
Wait 15 to 20 Minutes Before You Blow
The single biggest source of falsely high readings is mouth alcohol: residual vapor trapped in the mouth, throat, and esophagus that never came from your lungs. The device is designed to measure alcohol diffusing out of blood into deep lung air. Mix mouth alcohol into the sample and the result can be dramatically inflated.
Federal workplace testing rules require at least a 15-minute wait between a screening test and a confirmation test specifically to let mouth alcohol dissipate.2U.S. Department of Transportation. DOT Rule 49 CFR Part 40 Section 40.251 For a personal test, use the same 15-to-20-minute floor after your last sip of anything.
During the wait, avoid the things that put alcohol or alcohol-like compounds back in your mouth:
- Mouthwash and breath sprays, many of which contain significant alcohol.
- Liquid cough syrup and cold remedies, which often use alcohol as a solvent.
- Smoking and vaping, which can interfere with how the sensor reads the sample.
- Asthma inhalers with medications like albuterol, which some sensors mistake for alcohol.
- Food, and any burp or reflux episode, which push stomach gases into the mouth.
If you burp or feel reflux during the wait, restart the clock. Even a small amount of gas coming up from the stomach carries enough vapor to skew the reading.
How to Take the Test
Once the wait is done, the process itself is short but sensitive to technique. Errors in how you blow tend to push the number down or produce inconsistent results.
- Attach a fresh, unused mouthpiece. Residue from a previous sample contaminates the next one.
- Power the device on and let it warm up. Most units need 10 to 30 seconds for internal calibration and will indicate when they’re ready.
- Breathe in normally. Don’t hyperventilate or take an unusually deep breath first.
- Exhale with steady, continuous pressure into the mouthpiece. The reading stabilizes only when air from deep in the lungs reaches the sensor, so keep going until the device beeps or shows a done signal. Most require at least five seconds of continuous exhalation. Stopping early gives the sensor upper-airway air and understates your actual BAC.3National Center for Biotechnology Information. Alcohol Breath Testing
- Read the result, wait two to three minutes, and test again. If the two readings differ by more than 0.02, something went wrong with one of them; wait and retest.
A common mistake is puffing from the cheeks instead of exhaling from the lungs. Think about blowing out a candle held at arm’s length: sustained airflow, not a quick burst.
Reading the Number and What to Do With It
The display shows your BAC as a decimal percentage, meaning grams of alcohol per 100 milliliters of blood. In 49 states and Washington, D.C., the per se legal limit for driving is 0.08%.4Office of the Law Revision Counsel. 23 U.S. Code 163 – Safety Incentives to Prevent Operation of Motor Vehicles by Intoxicated Persons Utah sets it at 0.05%. Commercial drivers face a lower federal limit of 0.04%,5eCFR. 49 CFR 382.201 – Alcohol Concentration and every state enforces a zero-tolerance rule for drivers under 21, typically below 0.02%.
Here’s the part that matters most for how you use the number: your device has a margin of error. Even a quality consumer fuel cell can be off by 0.01 or more in either direction. Blowing 0.06 when the local limit is 0.08 does not give you a two-drink cushion. It gives you a reading inside the uncertainty range, and the only safe response is not to drive.
If you’re checking yourself over the course of an evening, run tests periodically and pay attention to whether the number is climbing or falling. A rising BAC means your body is still absorbing alcohol, and the reading you see now is lower than what you’ll peak at in another 30 to 60 minutes. Treat any detectable reading, not just one near the legal limit, as a reason to hand over the keys.
Why the Reading Can Still Be Off
Even with perfect technique and a good device, some sources of error can’t be eliminated by careful blowing.
The Partition Ratio
Every breathalyzer converts breath alcohol to blood alcohol using a fixed 2,100-to-1 ratio, assuming that 2,100 liters of deep lung air holds the same alcohol as one liter of blood. That ratio varies from person to person and even within one person over an evening. Research puts the actual average closer to 2,380 to 1, meaning a device calibrated to 2,100 tends to underestimate BAC by about 10 to 15 percent once your body is past absorption.6National Center for Biotechnology Information. Reflections on Variability in the Blood-Breath Ratio of Ethanol and Its Implications Higher body or breath temperature can push readings the other direction.
Testing Too Soon After a Drink
During the absorptive phase, alcohol is unevenly distributed and the concentration in your lungs can temporarily exceed the concentration in an arm blood draw, producing an overestimate. Absorption typically runs 30 to 90 minutes after your last drink, depending on whether you ate, how fast you drank, and your metabolism. The 15-to-20-minute wait is a minimum for clearing mouth alcohol, not a guarantee of a representative reading. For that, test at least an hour after your final drink.
Reflux and Diabetes
Two medical situations can produce readings that are simply wrong. Gastroesophageal reflux pushes alcohol-containing stomach gases up into the esophagus and mouth, and the device reads that regurgitated vapor as if it came from the lungs. If you have chronic reflux, expect unreliable numbers unless you can go a stretch without an episode. Separately, when the body produces ketones as fuel (from low blood sugar, insufficient insulin, or ketogenic states), one of them, acetone, is close enough to ethanol chemically that semiconductor sensors register it as alcohol. Fuel cells are much better at telling them apart but not perfectly immune. Someone in diabetic ketoacidosis can produce a positive reading with no alcohol at all.
Temperature and Batteries
Extreme cold or heat affects both the sensor’s chemistry and your breath. Testing right after stepping into winter air, or in intense heat, can shift the number. A low battery can also produce erratic results. If you see a low-battery warning, replace or recharge before testing.
Keep the Device Calibrated
An uncalibrated breathalyzer produces numbers, not measurements. Sensors drift with use and age.
For personal fuel cell devices, manufacturers typically recommend recalibration every 6 to 12 months or after 200 to 300 tests, whichever comes first. If accuracy matters to you, use the shorter interval. Most manufacturers offer mail-in calibration for $20 to $40. Skipping it turns a fuel cell into an expensive semiconductor.
Signs your device needs service:
- Readings that don’t match how much you’ve consumed.
- Three consecutive tests on the same breath that vary by more than 10 percent.
- Storage in a hot car, glove box, or freezing garage.
- You can’t remember the last calibration.
Professional evidential devices used in workplace programs follow a separate federal framework: manufacturers set the calibration intervals, and any device failing a calibration check must be pulled from service until repaired.7eCFR. 49 CFR 40.233 – Requirements for Proper Use and Care of EBTs
A Personal Breathalyzer Is Not a Roadside Test
One boundary worth stating: the reading on your consumer device isn’t the same instrument, legally or technically, as what police use. Roadside preliminary alcohol screening devices exist to help establish probable cause. Evidentiary breath tests happen on larger station-based machines that must meet National Highway Traffic Safety Administration standards, print results in triplicate, run air blanks between tests, and distinguish alcohol from acetone at the 0.02 level.8U.S. Department of Transportation. Approved Evidential Breath Measurement Devices Refusing an evidentiary test after arrest triggers implied consent penalties in every state, typically an automatic license suspension. Your personal reading is a private data point for your own decisions, nothing more.
Do the five things that eliminate the common errors: use a fuel cell device, keep it calibrated, wait at least 20 minutes after anything enters your mouth, blow steadily until the beep, and test twice. What’s left after that is physiological variation no consumer device can correct for, which is why a breathalyzer is an estimate and a blood draw is the reference standard.