Cannabis Just Passed Alcohol in Fatal Crash Toxicology. Nobody Has a Breathalyzer for It.
NHTSA spent years telling America that drunk driving was the impairment crisis. Turns out, the bigger number was growing behind it the whole time. A multi-site trauma center study across seven Level 1 facilities found cannabinoids in 25% of seriously or fatally injured road users, compared to 23% for alcohol.[1] Stimulants hit 11%, opioids 9%, and 32% of the fatalities had two or more substances in their system.
Two percentage points might not sound like a regime change, but it is. For decades, NHTSA built its entire enforcement apparatus around alcohol: breathalyzers, field sobriety tests, a universal .08 BAC threshold, implied consent laws. Every bit of that infrastructure targets one molecule, and cannabis has no equivalent. NHTSA itself has said it plainly: "There is no such measure for marijuana that correlates with impairment."[2] A driver with THC metabolites in their blood might have smoked an hour ago or a week ago; no test can tell you which.
NTSB flagged the same gap in its own safety research report, cataloging five systemic failures: inconsistent state drug laws, primitive roadside detection, no cannabis-specific countermeasures, unscreened prescription drugs, and an incomplete national crash data system.[3]
FARS data we've analyzed by vehicle shows where the shift is already visible at the model level. Across the database (2014–2023), drug-positive results appear in 9.6% of fatally crashed drivers versus 15.3% for alcohol. Those numbers lag the trauma center study because FARS drug testing depends on local jurisdiction and many coroners simply don't screen.[4] But specific models have already crossed the line: Suzuki's Grand Vitara at 12.7% drug-positive against 11.4% alcohol, Nissan's NV200 commercial van at 12.7% versus 11.8%, and Oldsmobile's Alero at a dead tie of 18.4% for both.[5]
NHTSA just launched "Drive High, Get a DUI" alongside its decades-old "Drive Sober or Get Pulled Over" campaign.[1] A slogan is not an enforcement tool. With 24 states and DC having legalized recreational cannabis since 2012, the roads got a new intoxicant with no reliable roadside detection. Breathalyzers took 30 years to become standard after their invention, and cannabis is moving faster than the science.
What You Can Do
If you use cannabis, treat it like alcohol and don't drive for several hours after edibles or smoking, because no validated threshold for safe THC levels exists. If you're buying a car, the vehicle doesn't change the chemistry, but FARS drug-positive crash rates vary by model by a factor of seven, from under 3% to over 18%, a spread driven by demographics rather than engineering. And if you're in a state that recently legalized, your roads just got a new variable and the enforcement toolkit hasn't caught up.
Sources & References
- NHTSA, Alcohol and Drug Prevalence Among Seriously or Fatally Injured Road Users, December 2025. Referenced in NHTSA’s “Drive Sober or Get Pulled Over” press release. nhtsa.gov
- NHTSA, Marijuana-Impaired Driving: A Report to Congress, DOT HS 812 440. nhtsa.gov
- NTSB, Alcohol, Other Drug, and Multiple Drug Use Among Drivers, Safety Research Report. ntsb.gov
- NHTSA, Research Note: Drug and Alcohol Crash Risk, DOT HS 812 355. nhtsa.gov
- NHTSA, Fatality Analysis Reporting System (FARS), 2014–2023. nhtsa.gov
Source: NHTSA FARS 2014–2023 for vehicle-level toxicology; NHTSA/NTSB trauma center study for national prevalence. FARS drug-positive rates are known undercounts due to inconsistent toxicology screening across jurisdictions. Cannabinoid-positive does not equal impairment at time of crash — THC metabolites persist for days to weeks, and no scientifically validated impairment threshold exists. See methodology for caveats.