Several factors can produce inaccurate breathalyzer readings, from residual alcohol in your mouth to the basic biology of how your blood carries alcohol. Some of these are within your control, others are quirks of your body chemistry, and a few come down to the device itself. Here’s what actually affects breathalyzer accuracy and how much each factor matters.
How Breathalyzers Detect Alcohol
Understanding what can go wrong starts with understanding how these devices work. There are two main types. Fuel cell breathalyzers use a platinum sensor that generates an electrical current when it reacts with ethanol in your breath. The stronger the current, the higher the reading. These are the handheld units police often use roadside. Evidential breathalyzers, the kind used at the station, typically use infrared spectroscopy, shining light at specific wavelengths through your breath sample to identify and measure ethanol molecules.
The key difference: fuel cell sensors react to the chemical itself but can’t always distinguish ethanol from similar compounds. Infrared devices analyze light absorption patterns at multiple wavelengths, which makes them better at filtering out non-ethanol substances. Both types, however, share a common vulnerability: they assume the alcohol in your breath came from your lungs, not your mouth.
Mouth Alcohol Is the Biggest Culprit
Breathalyzers are designed to measure “deep lung air,” the breath that has exchanged gases with your blood. When alcohol is sitting in your mouth or throat instead, the device reads it as if it came from your lungs and dramatically overestimates your blood alcohol. Fuel cell detectors are particularly susceptible because they cannot differentiate between mouth alcohol and alcohol carried from the lungs.
Common sources of mouth alcohol include mouthwash (Listerine and similar products contain significant amounts of ethanol), breath sprays, breath strips, cough syrups like NyQuil, cold medications, and topical oral products like Anbesol (a toothache gel with high alcohol content). Even large quantities of cough drops before a test can introduce enough residual alcohol to skew results.
This is why officers are typically required to observe you for 15 to 20 minutes before administering an evidential test. That waiting period allows mouth alcohol to dissipate. If that observation period is skipped or interrupted, any residual mouth alcohol can inflate the reading well beyond your actual blood alcohol concentration.
The Timing of Your Last Drink
Alcohol doesn’t hit your bloodstream the moment you swallow it. After your last drink, your body spends roughly 30 to 90 minutes absorbing alcohol into the blood, depending on factors like food in your stomach and how quickly you drank. During this “absorptive phase,” your BAC is still climbing.
If you’re tested while your BAC is still rising, the reading may be higher than it was when you were actually driving. Say you finish a drink, get in your car, drive for 15 minutes, and then get pulled over. By the time you blow into a breathalyzer 20 minutes later, your BAC could be meaningfully higher than it was behind the wheel. This lag between drinking and peak BAC is real and well-documented, and it’s the basis of what defense attorneys call the “rising BAC defense.”
Your Blood Chemistry Changes the Math
Every breathalyzer uses a fixed conversion factor to translate breath alcohol into blood alcohol: the 2100:1 partition ratio. This means the device assumes that 2,100 milliliters of your breath contains the same amount of alcohol as 1 milliliter of your blood. The problem is that this ratio varies from person to person.
One major variable is hematocrit, the proportion of your blood made up of red blood cells. Alcohol dissolves in plasma (the liquid part of blood), not in red blood cells. If you have a higher red blood cell count, you have proportionally less plasma per milliliter of blood. The same amount of alcohol gets concentrated into less liquid, which means more of it escapes into your breath. The breathalyzer reads this higher breath concentration and reports a BAC that’s higher than what a direct blood test would show. The reverse is true for people with lower hematocrit: their readings may come in lower than their actual blood alcohol. Breathalyzers don’t account for these individual differences at all.
Industrial Chemicals and Solvent Exposure
If you work around paints, adhesives, lacquers, or cleaning products, airborne chemicals can genuinely interfere with a breathalyzer. A documented case involving isopropanol (rubbing alcohol) exposure illustrates the problem well. An individual who had been exposed to isopropanol provided breath samples on two separate infrared instruments, which recorded apparent ethanol readings between 0.09 and 0.17 over roughly three hours. Later blood analysis confirmed the person’s actual ethanol level was only 0.076, with significant amounts of isopropanol and acetone also present in both breath and blood.
Modern infrared devices have built-in “interferant detectors” designed to flag non-ethanol substances, but these safeguards aren’t foolproof. Instrumental or procedural problems can still produce false-positive interferant results. Fuel cell devices, which lack this wavelength-based filtering, are even more vulnerable to chemical cross-reactivity.
Fermented Foods and Everyday Products
You may have heard that kombucha, ripe fruit, or fermented foods can trigger a false positive. The evidence suggests this concern is mostly overblown. A study testing regional fermented foods and medications found that none of the fermented foods produced false positives on breathalyzer tests. The only products that triggered false readings were cologne and a specific oral antiseptic, both of which contain high concentrations of alcohol that linger in the mouth.
The pattern is consistent: what matters isn’t trace ethanol in food sitting in your stomach, but alcohol-containing products that leave residue in your mouth and throat. A sip of kombucha with 0.5% alcohol by volume is unlikely to register. A gargle of mouthwash containing 26% alcohol, on the other hand, absolutely can.
What About Diabetes and Keto Diets?
This is one of the most persistent claims: that ketosis (from diabetes, fasting, or low-carb diets) produces enough acetone on the breath to fool a breathalyzer. The actual research tells a different story. A study that tested whether the ketone bodies produced during ketosis, including acetone, cross-reacted with breathalyzer technology found no false positives. At concentrations of 30 millimoles per liter (well above what you’d see in typical ketosis), acetone, acetoacetate, and beta-hydroxybutyrate did not trigger readings on either fuel cell or chemical detection methods.
Older semiconductor-based devices (the cheap personal breathalyzers) may be less discriminating, but the fuel cell and infrared instruments used by law enforcement appear to handle ketones without issue.
GERD and Acid Reflux
Another common claim is that gastroesophageal reflux disease (GERD) can push alcohol vapor from your stomach into your mouth during a breath test. The logic sounds plausible: a reflux episode could burp up stomach contents containing alcohol, contaminating the breath sample. However, a study specifically investigating this scenario concluded that the risk of stomach alcohol erupting into the mouth during a reflux episode and falsely increasing an evidential breath test result is “highly improbable.”
That said, if you had a major reflux episode immediately before or during a test, it’s not impossible for some stomach gas to reach the mouth. The 15-to-20-minute observation period is partly designed to catch exactly this kind of event.
Temperature and Calibration Issues
Breathalyzers are calibrated to operate at specific temperatures, typically around 68 to 70 degrees Fahrenheit. Extreme heat can cause sensor drift, affect battery performance, and expand internal components in ways that shift calibration. Rapid temperature swings, like moving a device from an air-conditioned patrol car into summer heat, can be particularly problematic.
Calibration itself is another potential weak point. NHTSA model specifications require that evidential devices maintain accuracy within 0.005 BAC at the 0.08 legal threshold. Devices are tested semi-annually or as necessary, and the manufacturer is responsible for ensuring proper calibration. If a department falls behind on calibration schedules, or if calibration records show the device was out of tolerance, the readings become legally and scientifically questionable. Fuel cell breathalyzers, in particular, have been noted to show greater variation from environmental conditions than infrared systems.
Body Temperature and Breathing Patterns
Your own body temperature plays a role too. The 2100:1 partition ratio assumes a normal body temperature of 98.6°F. A fever raises your core temperature, which increases the amount of alcohol that evaporates from your blood into your lung air. For every degree Fahrenheit above normal, breath alcohol readings can increase by roughly 6 to 8 percent. So a person with a 100.6°F fever could register a reading that’s about 12 to 16 percent higher than their actual BAC.
Hyperventilation and breathing patterns can also shift readings modestly in either direction. Hyperventilating before a test tends to lower the reading by clearing alcohol from the airways, while holding your breath can slightly increase it by allowing more alcohol to accumulate in the lung air. Officers are trained to watch for these behaviors, and most devices flag abnormal breath patterns.