- Cannabis impairs driving reaction time by 21% in meta-analysis studies (Hartman 2015), with peak impairment within 30–90 minutes of smoking — impairment largely resolves within 3–4 hours for occasional users.
- Unlike alcohol, THC blood levels do not reliably correlate with driving impairment — experienced users show near-normal driving performance at 5 ng/mL blood THC levels that cause significant impairment in occasional users.
- Washington and Colorado set per se DUI limits of 5 ng/mL THC in blood — this threshold was established without evidence it represents impairment and remains scientifically controversial.
- Most states use impairment-based DUI standards for cannabis — officers use field sobriety tests (SFST), Drug Recognition Expert (DRE) evaluation, and blood/urine tests to establish impairment.
- Cannabis’s effect on driving is dose-dependent: low doses (5mg THC) cause minimal measurable impairment; doses above 10–15mg impair hazard detection, lane tracking, and following distance judgment.
- Combining cannabis with alcohol produces additive impairment significantly worse than either substance alone — even low doses of both together approach the driving risk of alcohol alone at the legal limit.
- Employers with drug testing programs can terminate drivers for off-duty cannabis use due to DOT regulations — commercial vehicle operators (CDL) are zero-tolerance under federal law regardless of state legalization.
How THC Impairs Driving
THC binds to CB1 cannabinoid receptors concentrated in the basal ganglia, cerebellum, prefrontal cortex, and hippocampus — brain regions responsible for motor coordination, reaction speed, executive function, and time perception. The resulting impairment affects multiple cognitive domains simultaneously, each with direct implications for driving safety.
Reaction time: The most consistently documented impairment. Hartman and Huestis (2013) meta-analysis of 66 studies found a mean reaction time increase of 21% at typical recreational doses. At 60 mph, a 21% slower reaction time means a vehicle travels approximately 19 additional feet before the driver initiates braking — a distance that determines whether a collision with a stopped vehicle or pedestrian is avoided or not.
Hazard detection (divided attention): Cannabis impairs the ability to simultaneously monitor multiple inputs — a required skill for driving. THC-intoxicated drivers demonstrate “tunneling,” narrowing focus to one primary task while peripheral monitoring deteriorates. This is particularly dangerous because drivers may feel they are driving normally while missing hazards in their peripheral field.
Lane tracking (SDLP): Standard Deviation of Lateral Position, the standard measure of lane weaving in driving simulation research, increases significantly with acute THC. The effect is less pronounced than heavy alcohol intoxication but statistically significant at doses of 10mg THC and above. Lane-keeping impairment is dose-dependent and peaks at 45–90 minutes post-inhalation for smoked cannabis.
Speed regulation and following distance: Cannabis users under impairment tend to drive slower and leave larger following distances — a compensatory behavior that reflects some degree of self-awareness of impairment. However, this compensation is incomplete: hazard response deficits persist even when drivers slow down, and the compensation itself disappears at higher doses.
Duration of impairment by dose: Low dose (5mg THC smoked): measurable impairment peaks at 30–60 minutes, largely resolved by 2–3 hours. Moderate dose (10–15mg): peak impairment at 45–90 minutes, resolved by 3–4 hours for occasional users. High dose (20mg+): impairment may persist 4–6 hours. Edibles: impairment onset delayed 30–90 minutes, peak at 2–4 hours, may persist 6–8 hours due to slower absorption and longer elimination.
State DUI Laws for Cannabis
Every US state prohibits driving under the influence of cannabis, but the legal frameworks are not uniform. The two primary approaches are per se limits (a fixed blood THC threshold) and impairment-based standards (requiring demonstrated impairment). A third category, zero-tolerance, makes any detectable THC or metabolite illegal regardless of impairment or amount.
| State | Legal Standard | Blood THC Limit | Urine THC | Notes |
|---|---|---|---|---|
| Washington | Per se | 5 ng/mL whole blood | Not used for per se | First state to set 5 ng/mL; scientifically contested |
| Colorado | Per se (presumptive) | 5 ng/mL whole blood | Not used for per se | Presumption of impairment, rebuttable by defense |
| Montana | Per se | 5 ng/mL whole blood | Not used for per se | Recreational legal state |
| Nevada | Per se | 2 ng/mL whole blood | 15 ng/mL metabolite | Lower threshold than WA/CO; metabolite limit very broad |
| Illinois | Per se | 5 ng/mL whole blood | 10 ng/mL | Medical patients get no exemption from per se limit |
| California | Impairment-based | No fixed limit | No fixed limit | Officer observation + DRE + blood evidence; no per se |
| Massachusetts | Impairment-based | No fixed limit | No fixed limit | SFST + DRE; blood evidence admissible but no threshold |
| Arizona | Zero-tolerance | Any detectable THC | Any metabolite | Medical card is NOT a defense to zero-tolerance DUID |
| Utah | Zero-tolerance | Any detectable THC | Any metabolite | Applies to all cannabinoids including CBD-adjacent metabolites |
Zero-tolerance states present special risk for medical cannabis patients, who may have detectable THC metabolites in urine weeks after their last use. In Arizona, this has resulted in DUID prosecutions of individuals who consumed cannabis legally days before driving and were not impaired at the time of the stop. Medical cards provide no exemption in these states. See cannabis laws by state for your jurisdiction.
Blood THC vs. Impairment: The Scientific Problem
The fundamental problem with per se THC limits is that blood THC concentration is a poor proxy for impairment — a relationship that is far more complex and individual than blood alcohol content (BAC) for alcohol.
Alcohol distributes uniformly in body water and maintains a reliable pharmacokinetic relationship with cognitive impairment: nearly everyone at 0.08% BAC shows significant driving impairment. THC is fat-soluble and its distribution, metabolism, and behavioral effects differ dramatically between individuals based on use frequency, body composition, and CYP2C9/CYP3A4 metabolic enzyme activity.
Tolerance divergence: An occasional user who has consumed cannabis once or twice may show acute impairment at 2 ng/mL blood THC. A daily cannabis user may have baseline blood THC above 5 ng/mL upon waking — before their first consumption of the day — and show no measurable driving impairment at that level. Studies of heavy users have documented blood THC levels of 15–25 ng/mL with driving simulation performance indistinguishable from sober controls.
Time lag between blood THC and impairment: After smoking cannabis, blood THC peaks within 3–10 minutes and then drops rapidly. Peak cognitive impairment, however, occurs at 30–90 minutes — when blood THC has already fallen significantly. This means a blood draw taken 2 hours after consumption may show 2 ng/mL at a time when the driver was actually most impaired at the 1-hour mark with 8 ng/mL blood THC. The time-course disconnect is fundamental and cannot be resolved by adjusting the threshold.
The Colorado Supreme Court recognized the weakness of the 5 ng/mL threshold in People v. Brice (2019), allowing expert testimony that blood THC does not establish impairment. Multiple NHTSA-commissioned reviews have concluded that per se THC limits cannot be scientifically validated as impairment thresholds. Despite this, per se limits remain law in multiple states because they provide prosecutorial clarity that impairment-based standards lack.
DRE Evaluations and Field Testing
In states without per se limits — and as supporting evidence in per se states — law enforcement relies on Drug Recognition Expert (DRE) evaluation and Standardized Field Sobriety Tests (SFSTs) to establish impairment.
SFST validity for cannabis: The three standard SFSTs (Horizontal Gaze Nystagmus, Walk-and-Turn, One-Leg Stand) were developed and validated for alcohol impairment. Their validity for cannabis impairment is limited. HGN, the most reliable SFST for alcohol, shows no nystagmus with cannabis (THC does not cause the jerking eye movement alcohol does). W&T and OLS detect balance impairment, which is less prominent in cannabis DUI than in alcohol DUI. Studies show SFSTs have approximately 50–70% accuracy for identifying cannabis impairment — far below their 88%+ accuracy for alcohol at 0.08% BAC.
DRE 12-step protocol: The Drug Recognition Expert evaluation is a standardized 12-step assessment that includes breath alcohol testing, pulse, blood pressure, pupil size and reaction, divided attention tests, muscle tone assessment, injection site inspection, and interview. For cannabis, DREs look for: pulse above 90 bpm, conjunctival redness, body temperature elevation, eyelid tremor, lack of nystagmus (distinguishes cannabis from alcohol), green coating on tongue, and elevated blood pressure. DRE classifications correctly identify cannabis impairment approximately 80–85% of the time in controlled studies, though real-world accuracy is lower.
Roadside oral fluid testing: Several devices are approved in various states for roadside saliva THC testing. The Dräger DrugTest 5000 and Abbott SoTox are the most widely deployed. These devices detect THC above a threshold (typically 25 ng/mL in oral fluid) and can produce results in 8–10 minutes. Accuracy data from the Victoria Police (Australia) study showed 84% sensitivity and 95% specificity for THC at the 25 ng/mL threshold. Limitations include false negatives (oral fluid THC drops to below threshold rapidly), susceptibility to contamination from recent smoke inhalation, and invalid results in cold or humid conditions. Oral fluid THC is detectable for 1–24 hours after smoking, compared to 1–4 weeks for urine metabolites — making it a more impairment-relevant test than urine.
Cannabis + Alcohol: Multiplied Risk
The combination of cannabis and alcohol is the highest-risk scenario for driving impairment and represents a disproportionate share of cannabis-related traffic fatalities. The interaction is not simply additive — at sub-threshold doses of both substances, the combined impairment can exceed what either substance would produce at its own legal limit.
Hartman et al. (2015) demonstrated in a controlled driving simulator study that subjects who consumed both alcohol (to 0.065% BAC, below the legal limit) and cannabis (11.5 mcg/kg THC) showed driving impairment equivalent to or exceeding alcohol alone at 0.08% BAC. Neither substance alone at those doses produced DUI-equivalent impairment, but the combination did. This synergistic interaction is believed to occur because alcohol increases THC plasma levels by approximately 25% through slowing gastric emptying and altering intestinal absorption — meaning the same cannabis dose produces higher THC exposure when alcohol is present.
Data from roadside crash studies consistently find cannabis+alcohol combinations overrepresented in serious and fatal crashes relative to either substance alone. The National Roadside Survey found simultaneous cannabis and alcohol use in 7% of weekend nighttime drivers — a small fraction of all drivers but responsible for a disproportionately large share of alcohol-positive crash fatalities.
Legal implications: In states with per se alcohol limits, a driver may have both 0.06% BAC and 3 ng/mL blood THC — below both per se thresholds individually — yet face DUID prosecution under an impairment-based theory (combined impairment). Prosecutors in California and other impairment-based states routinely argue combination impairment in cases where neither substance level alone would support a per se charge.
Legal Consequences of Cannabis DUI
The legal consequences of a cannabis DUI conviction are substantial and extend well beyond the immediate fine and license suspension. Understanding the full penalty structure is important for risk assessment.
| Consequence | First Offense (Typical) | Second Offense (Typical) | Notes |
|---|---|---|---|
| Fine | $390–$1,800 (varies by state) | $1,000–$4,000 | Plus court costs, assessment fees, DUI school fees |
| License suspension | 90 days–1 year | 1–3 years | Some states allow restricted license for work purposes |
| Jail / custody | 0–6 months (often suspended first offense) | 10 days–1 year | Mandatory minimums apply in some states |
| Probation | 1–3 years | 3–5 years | Drug testing, treatment programs may be required |
| Insurance rate increase | 80–200% premium increase | Policy cancellation common | SR-22 filing required in most states for 3 years |
| Criminal record | Misdemeanor (most states, first offense) | Misdemeanor or felony | Felony if injury, child in vehicle, or prior DUI felony |
CDL (Commercial Driver’s License) holders: Commercial drivers are subject to federal DOT drug and alcohol testing regulations, which apply federal standards regardless of state law. Any detectable marijuana is a violation under DOT rules. A first CDL cannabis violation results in disqualification from commercial driving for a minimum of 1 year; a second violation is a lifetime disqualification. Off-duty cannabis use that produces a positive drug test will end a commercial driving career even in fully legal recreational states.
Immigration consequences: Non-citizens should be aware that cannabis DUI convictions can constitute a crime involving moral turpitude or a controlled substance offense under immigration law. This can affect visa applications, green card renewals, and naturalization proceedings. Non-citizens facing cannabis DUI charges should consult an immigration attorney in addition to a criminal defense attorney before accepting any plea agreement.
For state-specific legal information, see the cannabis legalization by state guide. For general cannabis safety information, see the effects guide and drug test guide.
Frequently Asked Questions
Yes, in all 50 states. DUID laws apply regardless of whether cannabis is recreationally or medically legal in your state. Legal standard varies: per se THC limits (5 ng/mL in WA/CO), impairment-based standards (CA, FL, TX), or zero-tolerance (AZ, UT). Legalization of cannabis does not exempt drivers from DUID prosecution.
For occasional users at moderate doses (5–10mg THC smoked): impairment resolves within 3–4 hours. A conservative safe guideline is 4–6 hours. For edibles: 6–8 hours minimum. For frequent users: shorter impairment window, but blood THC stays elevated longer. When in doubt, do not drive.
It varies by state. Washington, Colorado, Montana: 5 ng/mL whole blood. Nevada: 2 ng/mL blood / 15 ng/mL metabolite (urine). Illinois: 5 ng/mL blood / 10 ng/mL urine. Most states use impairment-based standards with no fixed limit. Zero-tolerance states (AZ, UT, and others) prohibit any detectable THC or metabolite.
You can refuse in most states, but implied consent laws treat refusal as equivalent to a failed test for licensing purposes — typically an automatic 1–3 year license suspension, often longer than the suspension for a first DUI conviction. Refusal can also be used as evidence of consciousness of guilt. Officers can obtain a warrant for a compelled blood draw in most states within hours of a stop.