Cannabis DUI Laws: How Impairment Is Tested and What the Penalties Are

Policy / Culture

No Valid Threshold

No scientifically validated THC blood threshold reliably indicates driving impairment, yet six states enforce a 5 ng/mL per se limit and others treat any detectable THC as automatic proof of intoxication.

Congressional Research Service, 2021

Congressional Research Service, 2021

View as image

Driving under the influence of cannabis is illegal in every state, regardless of whether marijuana is legal for recreational or medical use. That much is straightforward. What gets complicated, and what makes cannabis DUI fundamentally different from alcohol DUI, is the question of how impairment is detected, measured, and proven. The tools and standards used for alcohol do not translate cleanly to cannabis, and this gap between law enforcement methods and actual science creates real problems for both road safety and individual rights.

The Core Problem: Measuring Cannabis Impairment

Policy / Culture

Cannabis DUI: Testing Methods & Their Limitations

Blood THC Level

Measures: Active THC (ng/mL)

Accuracy: Poor — no reliable impairment threshold

Detection window: Hours to days (daily users: weeks)

Field Sobriety Tests

Measures: Motor coordination, balance

Accuracy: Low — designed for alcohol, not cannabis

Detection window: Current moment only

Drug Recognition Expert (DRE)

Measures: 12-step evaluation protocol

Accuracy: Moderate — high false-positive rate

Detection window: Current moment (subjective)

Oral Fluid Test

Measures: Recent THC in saliva

Accuracy: Better for recent use, still limited

Detection window: 2-24 hours since last use

State Legal Approaches

Per se limit (5 ng/mL)

CO, WA, MT, OH, NVOver limit = automatically guilty

Zero tolerance

AZ, GA, DE, IA, IN, othersAny detectable THC = DUI

Impairment-based

CA, NY, most statesMust prove actual impairment

The core problem: Unlike alcohol (BAC 0.08 = impaired), no blood THC level reliably predicts driving impairment. A daily medical patient may have 15 ng/mL THC and drive normally, while an occasional user at 3 ng/mL may be significantly impaired.

Clinical Chemistry, 2021 • Not legal adviceCannabis DUI Testing Methods

Alcohol impairment testing works because there is a direct, well-established relationship between blood alcohol concentration and impairment. At 0.08 percent BAC, the legal limit in all 50 states, virtually everyone shows measurable cognitive and motor deficits. The breathalyzer measures current blood alcohol levels with reasonable accuracy, and the relationship between that number and impairment is supported by decades of research.

Cannabis does not work this way. THC, the primary psychoactive compound, is lipophilic, meaning it is absorbed into fat tissue and released back into the bloodstream over time. A person's blood THC level does not correlate reliably with their level of impairment. Someone who consumed cannabis two hours ago may have declining blood THC levels but still be impaired. A daily medical patient may have elevated baseline THC levels but function normally because of developed tolerance.

Research published in journals including Clinical Chemistry and the Journal of Analytical Toxicology has repeatedly demonstrated that blood THC concentrations are poor predictors of driving performance. A 2021 meta-analysis by the Congressional Research Service concluded that while cannabis use increases crash risk by a modest factor, no specific THC threshold has been validated as a reliable indicator of impairment comparable to the 0.08 BAC standard for alcohol.

This is the fundamental scientific problem that cannabis DUI law has not solved.

Per Se THC Limits by State

Despite the scientific limitations, many states have enacted per se THC limits, meaning that driving with THC in your blood above a specified threshold is automatically illegal, regardless of whether you show any signs of impairment.

5 nanograms per milliliter states. Washington and Colorado, two of the earliest recreational legalization states, both set their per se limit at 5 ng/mL of active THC (delta-9-THC) in blood. Montana and Nevada use the same threshold. This number was chosen partly based on limited research suggesting impairment is more likely above this level, but the scientific community has not endorsed it as a reliable impairment marker.

Zero-tolerance states. Arizona, Georgia, Delaware, Indiana, Iowa, Oklahoma, Rhode Island, South Dakota, Utah, and Wisconsin have zero-tolerance or near-zero-tolerance policies. In these states, any detectable amount of THC or its metabolites in your blood can result in a DUI charge. This is particularly problematic because THC metabolites, especially THC-COOH, can remain detectable for weeks after last use. A medical patient in Arizona could face DUI charges while completely sober based on metabolites from cannabis consumed days or weeks earlier.

Effect-based states. Many states, including California, do not have a per se THC limit. Instead, they use an effect-based approach where the prosecution must prove the driver was actually impaired by cannabis at the time of driving. This requires evidence beyond a blood test, typically including officer observations, field sobriety test results, and sometimes Drug Recognition Expert evaluations.

Combination approaches. Some states use a hybrid model where a THC blood level above a certain threshold creates a rebuttable presumption of impairment that the defendant can challenge with evidence of tolerance or other factors.

Field Sobriety Tests and Cannabis

The Standardized Field Sobriety Test battery, which includes the horizontal gaze nystagmus test, the walk-and-turn test, and the one-leg stand test, was developed by the National Highway Traffic Safety Administration specifically for detecting alcohol impairment. These tests exploit the specific ways alcohol affects the body: disrupted smooth eye tracking, impaired balance, and reduced ability to perform divided-attention tasks.

Cannabis affects the body differently. THC does not typically cause horizontal gaze nystagmus, which is the involuntary jerking of the eyes that is the most reliable indicator in alcohol SFST testing. Cannabis can affect balance and coordination, but the manifestation is different from alcohol's effects, and the SFST scoring criteria were not calibrated for these differences.

A 2020 study published in the journal Psychopharmacology found that SFST performance was not a reliable predictor of cannabis-related driving impairment and that many sober individuals failed SFST components due to medical conditions, fatigue, anxiety, or physical limitations unrelated to drug use. The study noted that officers' assessments of impairment during SFSTs correlated more strongly with the officers' expectations than with actual THC levels or measured performance deficits.

Despite these limitations, SFST results remain admissible evidence in cannabis DUI cases in most jurisdictions. Defense attorneys can and do challenge their reliability, but juries may still find officer testimony about a failed field test persuasive.

Drug Recognition Expert Evaluations

When an officer suspects drug impairment but a breathalyzer shows no alcohol, the next step in many jurisdictions is a Drug Recognition Expert evaluation. DREs are officers who have completed a specialized training program administered through the International Association of Chiefs of Police.

The DRE protocol involves 12 steps: a breath alcohol test, an interview with the arresting officer, a preliminary examination and first pulse reading, eye examinations, divided-attention tests, vital sign checks, dark room examinations of pupil size, muscle tone assessment, injection site examination, statements and interrogation, the DRE's opinion, and a toxicological examination.

The DRE is trained to identify seven categories of drugs based on patterns of signs and symptoms. For cannabis, the expected indicators include elevated pulse, dilated pupils in certain lighting conditions, reddened conjunctiva, eyelid tremors, impaired time estimation, and reduced muscle tone.

The reliability of DRE evaluations is contested. NHTSA-funded studies have claimed accuracy rates of around 80 percent, but independent research has found lower reliability. A study in the Journal of Forensic Sciences found that DRE officers correctly identified cannabis as the impairing substance only about 44 percent of the time and produced false-positive identifications at significant rates. The subjectivity inherent in the evaluation, where the DRE must interpret physical signs that can have multiple causes, introduces unavoidable variability.

DRE evaluations carry significant weight in court, but they are opinions, not objective measurements. Effective legal defense often involves challenging the DRE's training, the specific observations made, and whether the physical signs observed could have alternative explanations.

Blood Testing: What It Shows and What It Does Not

If a driver is arrested for suspected cannabis DUI, a blood draw is typically requested or, depending on the jurisdiction, required under implied consent laws. The blood sample is tested for delta-9-THC, the active compound, and often for THC-COOH, the primary inactive metabolite.

Active THC in blood peaks within minutes of smoking and drops rapidly, often falling below 5 ng/mL within two to three hours for occasional users. However, for regular daily users, baseline blood THC levels can remain above 5 ng/mL for days after cessation because THC stored in fat tissue continually releases back into the bloodstream. A 2017 study in Clinical Chemistry found that some daily cannabis users had blood THC levels above 5 ng/mL after more than 24 hours of monitored abstinence.

The metabolite THC-COOH is even more persistent. It can be detected in blood for up to a week in occasional users and several weeks in daily users. States that test for metabolites rather than active THC are essentially testing for past use rather than current impairment.

The timing between a traffic stop and blood collection matters enormously. It typically takes one to three hours from the initial stop to the blood draw. During this window, THC levels can change significantly, rising in some cases as THC redistributes from tissues into blood, or falling as the body continues metabolizing the compound. This delay introduces additional uncertainty into an already imprecise measurement.

Penalties for Cannabis DUI

Penalties vary by state and by whether it is a first or repeat offense, but the consequences are serious across the board.

First offense penalties in most states include license suspension of 90 days to one year, fines ranging from $500 to $2,000, possible jail time of up to six months, mandatory substance abuse evaluation and education programs, probation of one to two years, and potential installation of an ignition interlock device. In some states, a first-offense cannabis DUI is a misdemeanor.

Second and subsequent offenses carry escalating penalties including longer license suspensions, higher fines, mandatory minimum jail sentences, extended probation periods, and in some states felony charges. A third DUI offense is a felony in most jurisdictions regardless of the substance involved.

Aggravating factors that increase penalties include having a minor in the vehicle, causing an accident, causing injury or death, having an extremely high THC level in per se states, and having a prior DUI of any kind on your record.

Collateral consequences beyond the criminal penalties can be equally significant. A DUI conviction appears on background checks and can affect employment, professional licensing, insurance rates, housing applications, and immigration status. For commercial driver's license holders, any DUI results in automatic CDL disqualification.

Emerging Testing Technologies

The inadequacy of current testing methods has driven significant investment in developing better tools. Several approaches are in various stages of development.

Oral fluid roadside testing devices analyze saliva for THC and are used in some jurisdictions in Canada and Australia. The Draeger DrugTest 5000 and Abbott SoToxa are the most widely deployed devices. These test for the presence of THC above a threshold, not for a specific concentration. While they can indicate recent cannabis use, they share the fundamental limitation of not measuring impairment. Several US states have piloted oral fluid testing programs.

Cognitive and psychomotor testing apps measure reaction time, divided attention, and other cognitive functions through tablet-based assessments. Companies including Druid and IMMAD have developed apps intended to provide objective measurement of impairment regardless of the substance causing it. These tools are promising in concept but have not yet been widely validated or adopted by law enforcement.

THC breath testing is being developed by companies like Hound Labs. The concept mirrors the alcohol breathalyzer but tests for THC in breath, which may correlate more closely with recent use than blood testing. As of 2026, these devices are in limited pilot use but have not achieved widespread deployment.

Practical Considerations for Cannabis Users

The safest approach is straightforward: do not drive after using cannabis. Research consistently shows that cannabis increases crash risk, with most studies estimating the increase at roughly 1.3 to 2 times baseline risk. This is lower than the increase associated with alcohol impairment, which ranges from 5 to 20 times at illegal BAC levels, but it is real and measurable.

If you use cannabis regularly, understand your state's specific laws. Know whether your state has a per se limit, what that limit is, and whether it tests for active THC or metabolites. If you are a medical patient in a zero-tolerance state, you face significant risk of a DUI charge even when completely unimpaired, and you should discuss this with your recommending physician and a defense attorney familiar with cannabis law.

If you are pulled over and an officer suspects impairment, know your rights regarding field sobriety tests and blood draws. In many states, you can refuse a field sobriety test without automatic penalty, though refusing a blood draw under implied consent laws typically results in automatic license suspension. An attorney experienced in cannabis DUI defense in your state can advise on the specific implications.

This article is for informational purposes only and does not constitute legal advice. DUI laws and enforcement vary by state and locality. Consult an attorney licensed in your jurisdiction for guidance on your specific situation.

The Bottom Line

Cannabis DUI law covering testing science, state approaches, and penalties. Core problem: blood THC levels do not correlate reliably with impairment (lipophilic, stored in fat, released over time); Clinical Chemistry and J Analytical Toxicology research confirms; 2021 Congressional Research Service meta-analysis — no validated THC impairment threshold comparable to 0.08 BAC. State approaches: per se 5ng/mL (WA, CO, MT, NV — not scientifically endorsed as impairment marker); zero-tolerance (AZ, GA, DE, IN, IA, OK, RI, SD, UT, WI — any THC/metabolites = DUI, problematic for medical patients); effect-based (CA and others — must prove actual impairment); hybrid/rebuttable presumption. Field sobriety: SFST battery (HGN, walk-and-turn, one-leg-stand) designed/validated for alcohol only; THC does not cause HGN; 2020 Psychopharmacology study — SFST not reliable for cannabis impairment, officer assessments correlated with expectations more than THC levels. DRE: 12-step protocol, IACP-trained; expected cannabis indicators (elevated pulse, dilated pupils, reddened conjunctiva, eyelid tremors, reduced muscle tone); J Forensic Sciences — DREs correctly identified cannabis only ~44% of the time; significant false-positive rates; opinions not objective measurements. Blood testing: active THC peaks minutes after smoking, drops below 5ng/mL within 2-3h for occasional users; 2017 Clin Chemistry — daily users above 5ng/mL after 24h+ abstinence; 1-3h delay between stop and blood draw introduces additional uncertainty. Penalties: first offense = license suspension 90d-1yr, $500-2K fines, up to 6mo jail, mandatory evaluation; escalating for repeats; third = felony most jurisdictions. Emerging tech: oral fluid roadside (Draeger DrugTest 5000, Abbott SoToxa), cognitive apps (Druid, IMMAD), THC breath testing (Hound Labs) — all limited deployment. Cannabis increases crash risk ~1.3-2x baseline (vs alcohol 5-20x at illegal BAC).

Frequently Asked Questions

Can you get a DUI for cannabis even if you are not impaired?

Yes, in many states. States with per se THC limits (like Colorado and Washington at 5 nanograms per milliliter) or zero-tolerance policies (like Arizona, Georgia, and Utah) can charge you with DUI based solely on blood THC levels, regardless of whether you show any signs of impairment. This is particularly problematic for regular or medical cannabis users, whose baseline blood THC levels can remain above per se limits for days or even weeks after cessation because THC stored in fat tissue continually releases into the bloodstream. A 2017 study in Clinical Chemistry found daily users above 5 nanograms per milliliter after more than 24 hours of monitored abstinence. In effect-based states like California, the prosecution must prove actual impairment, which provides more protection.

How much does cannabis actually increase crash risk?

Research consistently shows that cannabis increases crash risk by a modest factor, with most studies estimating the increase at roughly 1.3 to 2 times baseline risk. A 2021 Congressional Research Service meta-analysis confirmed this range. For context, alcohol impairment at illegal blood alcohol levels increases crash risk by 5 to 20 times, making it substantially more dangerous than cannabis for driving. However, the cannabis risk increase is real and measurable, and combining cannabis with alcohol produces a compounding effect that significantly exceeds either substance alone. The relatively modest risk increase does not make driving after cannabis use safe or legal — it remains illegal in every state and carries severe penalties.

Are field sobriety tests valid for detecting cannabis impairment?

The standard field sobriety test battery was designed and validated specifically for alcohol impairment and has not been scientifically validated for cannabis. THC does not typically cause horizontal gaze nystagmus, which is the most reliable indicator in alcohol SFST testing. A 2020 study in Psychopharmacology found that SFST performance was not a reliable predictor of cannabis-related driving impairment and that officers' assessments correlated more strongly with their expectations than with actual THC levels. Many sober individuals fail SFST components due to medical conditions, fatigue, anxiety, or physical limitations. Despite these limitations, SFST results remain admissible in most jurisdictions. In many states, you can refuse a field sobriety test without automatic penalty, though you should verify this in your specific state.

What is a Drug Recognition Expert and how reliable are they?

A Drug Recognition Expert is a specially trained police officer who conducts a 12-step evaluation protocol to identify drug impairment. The protocol includes breath alcohol testing, eye examinations, divided-attention tests, vital sign checks, pupil size assessment in dark conditions, muscle tone evaluation, and a toxicological exam. For cannabis, DREs look for elevated pulse, dilated pupils, reddened eyes, eyelid tremors, and reduced muscle tone. NHTSA-funded studies claim about 80 percent accuracy, but independent research in the Journal of Forensic Sciences found DREs correctly identified cannabis as the impairing substance only about 44 percent of the time, with significant false-positive rates. DRE evaluations are opinions based on subjective interpretation of physical signs that can have multiple causes. They carry weight in court but can be effectively challenged.

How long after using cannabis is it safe to drive?

There is no scientifically established safe waiting period that applies to everyone, because THC affects individuals differently based on tolerance, consumption method, and dose. General guidance from safety researchers suggests waiting at least 3 to 4 hours after smoking or vaping, and at least 6 to 8 hours after consuming edibles, before driving. However, these are rough guidelines, not guarantees. High doses, individual sensitivity, and combining cannabis with other substances can extend impairment beyond these windows. The most conservative and legally safe approach is to not drive at all on any day you have consumed cannabis. If you are a medical patient who uses cannabis daily, discuss driving safety explicitly with your recommending physician and understand your state's per se limits.

Can you refuse a blood test for cannabis DUI?

In most states, drivers have given "implied consent" to chemical testing by obtaining a driver's license. Refusing a blood test under implied consent laws typically results in automatic license suspension, often for a year or more, regardless of whether you were actually impaired. Some states treat refusal as an additional criminal charge. However, the Supreme Court ruled in Birchfield v. North Dakota (2016) that warrantless blood draws violate the Fourth Amendment, meaning officers generally need either your consent or a warrant to draw blood. In practice, officers can usually obtain a telephonic warrant quickly. You have the right to refuse a field sobriety test in many states without automatic penalty, but refusing chemical testing carries consequences. Consult an attorney experienced in cannabis DUI defense in your state for specific guidance.