Testing for marijuana DUI is a multi-step process that starts with a traffic stop and can escalate through roadside physical tests, a specialized officer evaluation, and eventually a blood or saliva sample sent to a lab. Unlike alcohol, there’s no quick, reliable breathalyzer equivalent for THC. That makes marijuana DUI cases more complex for both law enforcement and drivers.
What Happens at the Roadside
The process usually begins like any other traffic stop. If an officer suspects impairment, they’ll ask you to step out of the vehicle and perform field sobriety tests. These are the same physical coordination exercises used for alcohol: the Walk and Turn, the One Leg Stand, and sometimes additional tests like the Finger to Nose and a balance test called the Modified Romberg.
These tests pick up marijuana impairment, but not as cleanly as they detect alcohol. In a controlled study published in JAMA Psychiatry, officers correctly identified 81% of participants who had smoked THC as impaired. The catch is that they also flagged nearly half the placebo group as impaired, meaning the tests produce a lot of false positives. The One Leg Stand was the most useful single test: people who failed it were about three times more likely to show actual impairment on a driving simulator.
Officers also look for physical signs specific to cannabis. One key test checks whether your eyes can converge on a close object, like a pen tip moving toward your nose. About 69% of THC-dosed participants in that same study couldn’t converge their eyes properly, compared to 49% of the placebo group. Officers will also note red or bloodshot eyes, the smell of marijuana, delayed responses, and any observable driving errors that prompted the stop.
The Drug Recognition Expert Evaluation
If the officer believes you’re impaired but your breath alcohol test comes back low or zero, you may be evaluated by a Drug Recognition Expert, or DRE. These are specially trained officers who follow a standardized 12-step protocol developed by the International Association of Chiefs of Police and NHTSA. Not every department has a DRE on call, so this step doesn’t happen in every case.
The evaluation is thorough and takes roughly 45 minutes to an hour. After confirming that alcohol alone doesn’t explain the impairment, the DRE conducts detailed eye exams, including checking your pupil size in different lighting conditions during a dark room examination. They measure your blood pressure, pulse (taken three separate times during the evaluation), and body temperature. They check your muscle tone, since cannabis tends to cause relaxation rather than the rigidity seen with stimulants. They look for injection sites to rule out other drugs.
The DRE also repeats the divided attention tests from the roadside stop in a more controlled setting. After compiling all of this, the officer forms an opinion about which category of drug is causing impairment and requests a toxicological sample, usually blood, to confirm that opinion in a lab.
Chemical Testing: Blood, Saliva, and Urine
The chemical test is the piece of evidence that carries the most weight in court. Blood draws are the most common method for marijuana DUI cases because they measure active THC (delta-9-tetrahydrocannabinol) in your system. Some jurisdictions also collect oral fluid (saliva) or urine, though each sample type has limitations.
Active THC in blood peaks almost immediately after smoking and drops quickly, often falling below detectable levels within several hours. One study found that no subjects had detectable blood THC 10 hours after oral cannabis use, and researchers recommended that at least 9 to 10 hours should pass after consumption before driving. Oral fluid testing has a shorter detection window and, in that same study, returned zero THC readings at every time point after oral dosing, making it a poor match for edible cannabis.
The major scientific problem with chemical testing is that THC blood levels don’t reliably reflect how impaired someone actually is. A National Institute of Justice study found that THC concentrations in blood, urine, and oral fluid did not correlate with cognitive or psychomotor impairment. Many participants showed significantly decreased functioning even when their fluid samples contained low THC levels. This is partly because THC is fat-soluble. Regular users can have measurable baseline THC in their blood without being actively impaired, while an occasional user might be significantly impaired at the same concentration.
Roadside Oral Fluid Screening Devices
Some states are piloting portable oral fluid testing devices that work like a preliminary breath test for drugs. Two devices currently in use are the Abbott SoToxa and the Dräger DrugTest 5000, both of which screen for six drug classes including cannabinoids, opiates, and cocaine. A saliva sample is collected on a swab, and results come back in minutes.
These devices are screening tools, not confirmation tests. A positive result helps establish probable cause for an arrest and a formal blood draw, but it won’t be the primary evidence in court. Think of it as the drug equivalent of a roadside breathalyzer: useful for the officer’s decision-making, but not the final word.
THC breathalyzers are also in development. Cannabix Technologies has partnered with AlcoPro, a major U.S. supplier of testing instruments, to market a marijuana breath test that captures breath samples for lab analysis. The technology is in pre-launch stages and not yet widely deployed.
Legal Limits Vary by State
There is no national standard for how much THC in your blood constitutes impairment. States handle this in three main ways.
- Per se limits: Six states (Illinois, Ohio, Pennsylvania, Montana, Nevada, and Washington) set specific legal thresholds ranging from 1 to 5 nanograms of THC per milliliter of blood. If your blood test comes back at or above that number, you’re legally impaired regardless of how you were actually driving.
- Reasonable inference laws: Colorado uses a 5 ng/mL threshold, but treats it as a rebuttable presumption rather than an automatic conviction. A jury can infer impairment, but you can present evidence that you weren’t actually impaired.
- Effect-based laws: Many states have no specific THC threshold and instead require the prosecution to prove you were actually impaired through the totality of evidence: officer observations, DRE evaluation, driving behavior, and chemical test results.
The science behind these thresholds is shaky. Research shows that blood THC at 5 ng/mL or above is associated with crash risk comparable to a blood alcohol concentration of 0.15%, which is nearly twice the legal alcohol limit. But the relationship isn’t linear or predictable the way it is with alcohol. A frequent user might function normally at 5 ng/mL while an occasional user could be dangerously impaired at 2 ng/mL.
What Happens If You Refuse Testing
Every state has implied consent laws, meaning that by driving on public roads you’ve already agreed to submit to chemical testing if lawfully arrested for impaired driving. Refusing a blood or fluid test after arrest triggers automatic penalties, typically a license suspension that can last a year or more. In many states the refusal itself can also be used as evidence of guilt at trial.
Refusing the roadside field sobriety tests is a different matter. In most states, these are voluntary before an arrest, and declining them doesn’t carry the same statutory penalties. However, an officer can still arrest you based on other observations, and your refusal may be noted in the police report.
Why Marijuana DUI Testing Remains Imperfect
The core challenge is that cannabis doesn’t behave like alcohol in the body. Alcohol produces a relatively clean dose-response curve: higher blood alcohol means more impairment, and the correlation is strong enough to anchor laws around 0.08%. THC doesn’t work that way. It’s absorbed into fat tissue, released unpredictably, and affects people differently based on tolerance, consumption method, and individual biology.
This means marijuana DUI cases rely more heavily on observed behavior and officer judgment than alcohol cases do. The field sobriety tests catch most impaired drivers but also flag a significant number of sober ones. The chemical tests confirm recent use but can’t precisely measure impairment. And the legal thresholds, where they exist, are based on population-level crash risk data rather than a clean biological marker of being too intoxicated to drive. If you’re pulled over and suspected of driving high, expect a process that’s longer, more subjective, and more dependent on the totality of evidence than a standard alcohol DUI stop.