Among nearly 4,000 crash victims in British Columbia and Ontario, THC was detected in 12.4%—roughly as common as alcohol—and drivers and passengers had similar THC rates, suggesting cannabis impairment laws aren't deterring use before driving.
Traffic safety researchers, policymakers designing cannabis-impaired driving laws, and anyone who drives after using cannabis.
What the researchers found
This study analyzed toxicology data from 3,004 drivers and 941 passengers who were moderately injured in motor vehicle accidents across British Columbia and Ontario. The researchers wanted to disentangle two effects: general cannabis availability (which would affect both drivers and passengers equally) from the specific deterrent effect of driving-under-the-influence laws (which should make drivers less likely to test positive than passengers).
Alcohol showed the expected pattern: passengers had higher detection rates than drivers (aPR 1.22), consistent with the idea that DUI laws deter some people from drinking before driving (while passengers face no such deterrent). But THC showed no such difference—drivers and passengers were equally likely to test positive, at a rate of 12.4%.
This absence of a driver-passenger gap for THC is concerning. It suggests that cannabis-impaired driving laws aren't creating the same deterrent effect as alcohol DUI laws. Possible explanations: people may not realize cannabis impairs driving, may not fear detection (no reliable roadside test), or may not believe they'll be caught.
The similar prevalence of THC (12.4%) and alcohol (14.2%) among crash victims is itself noteworthy—cannabis is now rivaling alcohol as a substance detected in crash-injured people.
Why it matters
If cannabis driving laws aren't deterring use before driving the way alcohol laws do, that has direct implications for road safety policy. The finding suggests current cannabis-impaired driving enforcement—hampered by the lack of a reliable roadside test (see RTHC-00173)—isn't achieving the behavioral change that decades of alcohol enforcement have produced.
The numbers in context
3,004 drivers + 941 passengers. Alcohol detection: 14.2%. THC detection: 12.4%. Passengers had higher alcohol than drivers (aPR 1.22). No driver-passenger difference for THC. 55.1% male, mean age 43.8.
How the study worked
Chart review and toxicology data from an ongoing prospective study of moderately injured motor vehicle occupants in BC and Ontario. 3,004 drivers and 941 passengers. Log-binomial regression models for prevalence ratios. Approximately 55.1% male, mean age 43.8 years.
Who was studied
N=3,004 drivers and 941 passengers, mean age 43.8 years, 55.1% male, from British Columbia and Ontario, Canada.
What this study cannot tell us
Toxicology was performed on moderately injured crash victims, not a random sample of drivers—the subset involved in crashes may differ from all drivers. THC detection indicates recent use but not necessarily impairment at the time of the crash. BC and Ontario may not represent all of Canada or other countries. The driver-passenger comparison assumes passengers face no deterrent, which may not be entirely true (e.g., some passengers become drivers later).
How to read the evidence
Prospective toxicology study of crash victims with driver-passenger comparison—strong real-world data though limited to moderately injured occupants.
When this study was published
Published in 2025 with data from BC and Ontario, reflecting post-legalization Canadian driving patterns.
The bigger picture
This directly extends RTHC-00143's finding that 54.9% of injured Canadian drivers tested positive for at least one substance. This study adds the driver-passenger comparison, which reveals the enforcement gap specific to cannabis. Combined with RTHC-00173's call for better THC pharmacokinetic data and RTHC-00171's workplace testing analysis, the picture is clear: current tools for detecting and deterring cannabis-impaired driving are inadequate compared to alcohol.
Replication
Not stated in abstract.
Funding
Not reported in abstract.
Conflicts of interest
Not reported in abstract.
Questions still open
- Would effective roadside THC testing create a deterrent effect similar to breathalyzers? Is the lack of driver-passenger difference unique to Canada or seen in other jurisdictions? Do cannabis users genuinely believe they can drive safely, or do they simply not fear enforcement?
Read the original research
A comparison of the prevalence of cannabis and alcohol use among drivers and passengers in British Columbia and Ontario, Canada.
Accident; analysis and prevention, 222, 108242
Accident; analysis and prevention is a reputable journal focusing on research related to traffic safety and accident prevention.
Citation
Pei, Lulu X; Chan, Herbert; Besserer, Floyd; Eppler, Jeffrey; Lee, Jacques; MacPherson, Andrew; McGrath, Michael; Ohle, Robert; Taylor, John; Vaillancourt, Christian; Brubacher, Jeffrey R. (2025). A comparison of the prevalence of cannabis and alcohol use among drivers and passengers in British Columbia and Ontario, Canada.. Accident; analysis and prevention, 222, 108242. https://doi.org/10.1016/j.aap.2025.108242