Marijuana impairs nearly every skill you need behind the wheel: reaction time, lane tracking, attention, short-term memory, and the ability to handle multiple tasks at once. The effects are dose-dependent, meaning more THC leads to worse performance, and they’re most pronounced in the first one to three hours after smoking. A single low-potency joint produces body sway and delayed braking comparable to driving at the legal alcohol limit in many countries (0.05% blood alcohol).
What THC Does to Driving Skills
Driving feels simple when it’s going well, but it actually demands constant coordination of perception, decision-making, and motor control. THC disrupts all of these. The most consistently measured impairment is slower reaction time, which showed up in four out of six controlled studies even at relatively low doses. That delay matters most in situations where you need to brake suddenly or respond to an unexpected hazard.
The other hallmark effect is lane weaving. Researchers measure this as the standard deviation of lateral position (SDLP), essentially how much your car drifts side to side within or outside your lane. In on-road tests, doses of roughly 14 and 21 mg of THC increased lane weaving by 1.7 to 3.5 centimeters compared to placebo. That may sound small, but it’s enough to push you into another lane or off a shoulder, especially at highway speeds.
Complex tasks are hit hardest. Simple driving on an empty road may feel manageable, but the moment you add a second demand, like following a lead car while watching for road signs, performance drops significantly. THC users in simulator studies increased their following distance and became more erratic in maintaining it, which signals difficulty processing multiple inputs at once. Visual search speed and the ability to hold new information in working memory also decline, particularly at higher doses.
How It Compares to Alcohol
Researchers have directly compared THC-related lane weaving to the effect of alcohol. In one well-designed simulator study, every 1 microgram per liter of THC in the blood increased lane weaving by about 0.26 cm, while every incremental rise in breath alcohol increased it by 0.42 cm. Translated into practical terms, a blood THC level of about 8 micrograms per liter produced the same degree of weaving as a 0.05% blood alcohol level, and 13 micrograms per liter matched a 0.08% blood alcohol level, the legal limit for alcohol in most U.S. states.
There’s an important behavioral difference, though. People who are high on marijuana often seem partially aware they’re impaired. They tend to slow down, increase following distance, and avoid risky passes. Alcohol does the opposite, reducing inhibition and encouraging overconfidence. This compensation likely explains why some crash statistics for marijuana look less dramatic than those for alcohol. But compensation has limits. It doesn’t help when a child runs into the road or traffic stops suddenly, situations where slowed reaction time and impaired attention can’t be offset by driving a few miles per hour slower.
Crash Risk by the Numbers
The crash risk picture is more complicated than you might expect. A large NHTSA case-control study initially found that THC-positive drivers had a 25% higher crash risk than sober drivers. But once researchers adjusted for age, gender, race, and alcohol use, the increased risk essentially disappeared, landing at an adjusted odds ratio of 1.00. In other words, the demographics of who tends to use marijuana (younger males, people who also drink) accounted for much of the raw risk.
Other analyses reach different conclusions. A widely cited meta-analysis found that cannabis use roughly doubled the risk of a motor vehicle collision (odds ratio of 1.92), with even higher risk in studies focused on fatal crashes (odds ratio of 2.10). The European DRUID project placed THC-positive drivers in a “slightly increased risk” category, with odds ratios between 1 and 3 times those of drug-free drivers. The discrepancy between studies largely comes down to methodology. Studies that can control for who uses marijuana and under what circumstances tend to find smaller effects, while simpler comparisons show larger ones.
What’s unambiguous is the danger of combining marijuana with alcohol. The two substances produce additive impairment on lane weaving. A blood THC level of 5 micrograms per liter combined with a 0.05% blood alcohol level produces the same degree of impairment as 0.08% blood alcohol alone. Fatalities involving both cannabis and alcohol doubled from about 4.8% of traffic deaths in 2000 to 10.3% in 2018.
Daily Users vs. Occasional Users
People who use cannabis daily do develop some tolerance to its impairing effects. In one study comparing daily and occasional users after smoking cannabis with 15 to 30% THC, occasional users showed a statistically significant increase in lane weaving while daily users did not, even though daily users had far higher blood THC levels (an average of 36 ng/mL vs. 6.4 ng/mL). Both groups reported a similar subjective high, rating it around 50 on a 100-point scale.
Daily users also showed a distinct behavioral pattern: they slowed down by about 1 mph on average after using cannabis, while occasional users and non-users maintained or slightly increased their speed. This suggests daily users may be more practiced at compensating. However, tolerance is partial, not complete. Complex cognitive tasks, the ones that matter most in real-world emergencies, are less susceptible to tolerance than simpler ones. And the fact that daily users feel equally high while showing less measurable impairment doesn’t mean they’re safe. It means their baseline is harder to judge, both for themselves and for law enforcement.
How Long the Impairment Lasts
For smoked or vaped cannabis, impairment peaks within the first 40 to 100 minutes. In on-road driving tests, THC-dominant cannabis significantly increased lane weaving during this window (by about 2.3 to 2.8 cm). By four to five hours after use, standard lane-tracking measures returned to placebo levels in most studies. However, more complex cognitive tasks, like dividing attention between multiple visual targets, still showed significant impairment up to five hours after use.
A conservative guideline suggested by clinical researchers is to wait at least five hours after using any THC-containing product before driving, and longer if you’ve also consumed alcohol. For people in jurisdictions with zero-tolerance THC laws, or those who want a wider safety margin, at least 12 hours is a more cautious target. Edibles present an additional complication: they take longer to kick in (often 60 to 90 minutes or more) and their effects last longer, which means the impairment window extends well beyond what you’d expect from smoking. Precise timelines for edibles are less well studied, but the delayed and prolonged peak makes them especially risky if you’re planning to drive later in the day.
Why There’s No Legal “THC Limit” Like Alcohol
Several states have set legal thresholds for THC in the blood, typically 5 ng/mL, but these numbers don’t work the way alcohol limits do. Blood alcohol concentration maps reliably onto impairment: at 0.08%, virtually everyone is measurably worse at driving. THC doesn’t follow that pattern. Blood THC levels drop rapidly after smoking, often falling below 5 ng/mL within one to two hours, while the person may still be significantly impaired. Meanwhile, a daily user could have a resting THC level well above 5 ng/mL while completely sober.
NHTSA has acknowledged this problem directly, stating that the poor correlation between blood THC levels and impairment makes it impossible for toxicologists to testify that a specific THC concentration means a driver was too impaired to drive safely. This is why there’s no national per se limit for THC the way there is for alcohol. Law enforcement currently relies on Drug Recognition Experts, trained officers who conduct a series of physical and cognitive evaluations at the roadside, along with oral fluid testing that can confirm recent cannabis use but can’t determine your level of impairment.
Oral fluid (saliva) tests can detect THC at concentrations as low as 2 ng/mL with confirmation testing, but they only tell you whether someone has used cannabis recently, not whether that person is impaired right now. This gap between detection and impairment remains one of the biggest challenges in cannabis traffic safety, and it’s a meaningful reason to build in your own safety margin rather than relying on a legal threshold that may not protect you.