Clicky

Cannabis

THC detection and impairment: Why a test result does not answer both questions

October 8, 2026
11 Min Read

By HealthDataConsortium.org Research Team

A positive THC result and a driver who is impaired by THC are two separate facts, and no single test result establishes both. Detection asks whether THC or one of its breakdown products is present in a specimen. Impairment asks whether a person’s driving-related skills are affected right now. The 2017 U.S. National Highway Traffic Safety Administration (NHTSA) report to Congress on marijuana-impaired driving concludes that, unlike alcohol, THC levels do not reliably track impairment. This guide defines both questions, shows where they split, and walks one illustrative roadside scenario through the evidence.

The Short Answer

Detection tells you that THC exposure happened at some point. Impairment tells you how a person is functioning at a particular moment. A test can answer the first question without answering the second, and a low or even negative reading does not rule out impairment either. NHTSA says the level of THC in blood or oral fluid does not appear to be an accurate and reliable predictor of impairment. Public health pages from the National Center for Complementary and Integrative Health (NCCIH) add that cannabis use has been linked to a higher risk of motor vehicle crashes. So the research supports taking impairment seriously while also showing that a lab number cannot measure it.

Two Questions, Two Kinds of Evidence

Detection: is something there?

Detection is a chemistry question. A laboratory looks for THC or a metabolite in blood, oral fluid, urine, hair or sweat and reports whether the amount reaches a set cutoff. How long a compound stays detectable depends on the compound, the specimen, the cutoff and the person’s pattern of use. Our guide to THC drug-test cutoffs and what a positive urine result means covers how cutoffs work, and this explainer on THC metabolites and active compounds separates the lab terms.

Impairment: is function affected?

Impairment is a performance question. NCCIH describes THC as the substance mainly responsible for marijuana’s effects on a person’s mental state, and the FDA describes it as the component that produces the “high.” NHTSA’s review of laboratory, simulator and instrumented-vehicle studies reports that marijuana can slow reaction time, increase lane-position variability, reduce divided attention and affect judgment. Those studies usually compare performance with and without a dose. They typically do not measure blood THC, so they show that impairment can happen without tying it to a particular number.

Where the Two Questions Diverge

  • Timing in the blood. NHTSA reports that after smoking, THC peaks at or just after the end of smoking and then drops quickly. By an hour after smoking, peak levels have declined by 80 to 90 percent, and after a few hours only low levels or none remain. In the Huestis review’s controlled smoking study, mean plasma THC was about 20 percent of its peak 30 minutes after smoking. Effects on driving-related skills are described as lasting hours, not days or weeks.
  • Peak level versus peak impairment. In a study NHTSA summarizes, peak performance deficits came about 90 minutes after smoking, when THC had already fallen by more than 80 percent from its peak. The subjective “high” did not line up with blood THC either.
  • Storage. THC is fat soluble. NHTSA notes it can be stored in fatty tissue and released back into the blood long after use, and that some studies have found it in blood 30 days afterward, even though the acute effects last hours.
  • Parent compound versus metabolite. A review by Marilyn Huestis in Chemistry & Biodiversity describes THC-COOH, the metabolite urine screens target, as inactive. In one controlled smoking study it stayed detectable in plasma for an average of 84 to 152 hours, depending on dose, while plasma THC was 5 ng/mL or lower within about two hours. The review states that a positive urine test shows only that exposure occurred, and not the route, the amount, the timing or the degree of impairment.
  • Specimen type. Oral fluid detects the parent compound THC for a shorter time than urine and is more indicative of recent use, according to the same review. It also cautions that wide differences between and within individuals make oral-fluid concentrations unreliable indicators of impairment.
  • The alcohol comparison. NHTSA contrasts THC with alcohol. Breath and blood alcohol concentrations rise and fall with impairment fairly consistently, which is why per se alcohol limits rest on evidence. NHTSA says THC has no equivalent chemical test or established impairment standard.

A Worked Comparison: One Roadside Scenario

This is a constructed illustration of how to read the evidence. It is not a real case, and it is not a prediction for anyone.

Imagine two different drivers are each stopped by an officer who suspects impairment. A blood sample is later drawn from each, and both samples show THC at a few nanograms per milliliter (ng/mL).

  • Driver A smoked cannabis within the previous hour or two.
  • Driver B uses cannabis regularly but has not used it recently.

The two blood results look alike. NHTSA says low THC levels of a few ng/mL can come from relatively recent use, when some slight or even moderate impairment is likely, or from chronic use with no recent ingestion and no impairment. The number alone cannot tell the two situations apart.

The timeline adds more uncertainty. NHTSA notes that blood is almost always drawn hours after use ended, because time passes between smoking, driving, the stop and the draw. In Washington State data it cites, the median time to a blood draw was about 165 minutes overall and about 139 minutes for drivers who tested positive for THC, and longer delays went with lower THC concentrations. A level measured at the draw therefore describes a different moment than the moment of driving.

Per se THC limits show the same problem at scale. NHTSA describes a Washington State laboratory study of drivers suspected of marijuana impairment whose blood tested above 1 ng/mL. About 24 percent were below a 2 ng/mL threshold, and about 63 percent were below a 5 ng/mL threshold. A separate Washington State analysis (Logan, Kacinko and Beirness, 2016) found that THC concentration correlated poorly with performance on roadside psychophysical tests. NHTSA concludes that such limits do not appear to rest on scientific evidence of impairment.

Now compare what different readings can and cannot support:

  • A blood THC number. It can show that THC was present when the sample was taken. It cannot, by itself, show how impaired the person was, or that they were not.
  • A urine screen positive at a cutoff. It can show past exposure. NHTSA says urine results cannot prove a driver was under the influence at the time and that detection of THC or other cannabinoids in urine does not necessarily reflect recent use.
  • An oral-fluid result. It suggests more recent exposure than urine does. NHTSA also says on-site oral-fluid screening devices have not been shown to be completely accurate and reliable, and that distinguishing use from environmental exposure, such as secondhand smoke, is not fully resolved.
  • Hair or sweat results. These can suggest past exposure over a longer period, but they cannot show impairment at the time of a stop or crash.
  • An officer’s observations. NHTSA notes that almost all of the signs and symptoms in the Drug Evaluation and Classification program are not based on driving impairment, and that no evidence-based method yet exists to detect marijuana-impaired driving or to separate it from alcohol impairment.

The practical takeaway is that each piece of evidence answers part of the question and none answers all of it.

Why Detection Windows Vary So Much

Detection windows are not one fixed number because several things change them. The pharmacokinetics review describes dose, route, rates of metabolism and excretion, assay sensitivity and the cutoff used. In the same review’s single-cigarette studies, the last positive urine sample at a 15 ng/mL cutoff averaged about 34 hours after a 1.75% THC cigarette and about 89 hours after a 3.55% THC cigarette. One heavy-user case tested positive by immunoassay (above 20 ng/mL) up to 67 days after last use. That spread, from a day or two to more than two months, is why a single detection window cannot be quoted for everyone. Most of the controlled research NHTSA relies on involves smoked cannabis, so the timelines may not carry over to edibles, which are absorbed more slowly and have effects that last longer. For the detection side of the story in more depth, see why THC can still show in urine after 30 days and how screening, confirmation and medical review work.

What the Research Does Say About Impairment and Risk

Impairment research is not a reason to dismiss concerns about driving after cannabis. The laboratory work NHTSA reviews finds measurable effects on skills that driving depends on, and NCCIH lists an increased risk of motor vehicle crashes among cannabis-related safety concerns. At the same time, NHTSA says the research on crash risk is mixed. Studies using different methods have produced contradictory estimates, and its own large case-control study in Virginia Beach found an unadjusted odds ratio of 1.25 for THC-positive drivers. After adjusting for age, gender and race or ethnicity it was 1.05 (95% confidence interval 0.86 to 1.27), and after also adjusting for alcohol it was 1.00 (0.83 to 1.22), meaning no measurable added crash risk in that study. The agency cautions that study designs differ and that more data are needed.

No source reviewed here gives a waiting time after which driving is known to be safe, and this article does not offer one. For how public health agencies frame that gap, see our guide to what research can and cannot say about cannabis and driving.

Questions to Ask When You See a Result

  1. Is the question about presence or about function? A lab number speaks to presence, and only an assessment of performance speaks to function.
  2. Which compound was measured: THC itself, or a metabolite such as THC-COOH?
  3. Which specimen was tested, and which cutoff applied?
  4. How much time passed between the event in question and the sample collection?
  5. Was a screening result confirmed by a second, more specific method?
  6. Is any other evidence being used, such as observed behavior or a structured evaluation, and who documented it?

What This Article Cannot Tell You

  • It cannot interpret your own result, estimate how long THC stays detectable for you, or say how a result will affect your job, license or legal situation.
  • It does not say when anyone can safely drive, and it gives no guidance on timing, preparing for or influencing a test.
  • Its main sources are a 2017 federal report and a 2007 review. NCCIH’s page was last updated in November 2019 and the FDA’s content is current as of March 2020. Newer studies may refine these details.
  • It is not medical or legal advice. It has not been clinically or peer reviewed. HealthDataConsortium.org is an independent publication whose contributors are researchers, not licensed clinicians. See the site’s Medical Disclaimer and Editorial Standards.

Who Should Get Professional Advice

If a test result or an impaired-driving question affects your employment, license or legal status, talk to a licensed attorney in your state or to the program’s reviewing physician. If you use cannabis for a health reason or take other medicines, ask your clinician or pharmacist about how it may affect you.

Related Reading on HealthDataConsortium.org

Sources Checked

This article contains no affiliate links and no product recommendations.

Related guides