After controlled THC dosing, cannabis users who said they felt safe to drive performed better on the simulator than those who didn't — but self-assessment missed some impairment, especially among experienced users.
Read this if you ever drive after using cannabis — your gut check about whether you're safe to drive is useful but not perfect.
What the researchers found
Can cannabis users tell when they shouldn't drive? This study takes a more granular approach than the French study (RTHC-00093) by examining not just whether self-assessment correlates with impairment, but what factors make self-assessment more or less accurate.
Regular cannabis users (at least monthly) were dosed with cannabis containing approximately 6.18% THC, then drove on a simulator at 30, 90, and 180 minutes post-dose. Before each drive, they answered a simple yes/no question: 'Do you feel safe to drive on real roads right now?'
The results showed self-assessed readiness to drive (RTD) had predictive value. Participants who said 'no, I don't feel safe to drive' showed more driving impairment (measured by standard deviation of lane position — SDLP) than those who said yes. Previous experience driving after cannabis use also mattered: people who had driven within 2 hours of cannabis use in the past, and those who did so more frequently, showed different performance patterns.
But self-assessment wasn't perfect. Some participants who said they felt safe to drive still showed impairment on the simulator. The gap between perceived and actual safety is the critical concern — it's the window where impaired driving occurs because the user doesn't recognize their impairment.
The study design parsed driving into specific events (lane changes, curves, intersections) rather than averaging across the whole drive, allowing a more detailed picture of where impairment manifests.
Why it matters
The practical question for millions of cannabis users isn't abstract — it's 'can I drive right now?' This study provides evidence that self-assessment has value but isn't foolproof. Users who say they feel unsafe to drive should absolutely not drive, but users who feel safe may still be impaired. This has direct implications for public health messaging: encouraging self-assessment as a first check is worthwhile, but it shouldn't replace objective standards.
The numbers in context
Cannabis dose: ~6.18% THC. Drives at 30, 90, and 180 minutes post-dose. RTD (readiness to drive) assessed before each drive. SDLP (standard deviation of lane position) measured within driving events. Participants who said they didn't feel safe to drive showed more impairment. Prior experience driving after cannabis use influenced both self-assessment and performance.
How the study worked
Observational study with controlled cannabis dosing. Participants who used cannabis at least monthly completed a baseline drive, were dosed with ~6.18% THC, then drove at ~30, 90, and 180 minutes post-dose. Self-reported readiness to drive (RTD, yes/no) assessed before each drive. Venous blood drawn at baseline and ~15 minutes post-dose. Cannabis use history obtained including prior experience driving within 2 hours of use. Driving segmented into events; SDLP measured within events.
Who was studied
N=30 individuals who use cannabis at least monthly, US participants
What this study cannot tell us
Relatively low THC dose (6.18%) compared to typical consumer products (often 20%+). Driving simulator, not real-world driving. Self-reported readiness was a simple yes/no — a scaled response might capture more nuance. Study population limited to regular users (at least monthly), so findings may not apply to occasional users who likely have less calibrated self-assessment. The segmented event analysis is innovative but complex to interpret.
How to read the evidence
Controlled dosing study with simulator assessment. Well-designed for measuring the self-assessment question, but limited by the low THC dose, simulator setting, and restriction to regular users.
When this study was published
Published in 2024. Cannabis and driving research continues to evolve with more sophisticated methods and higher-potency products.
The bigger picture
This builds directly on RTHC-00093 (French study finding cannabis users accurately self-assess impairment). Both studies agree self-assessment has value, but this study adds nuance: prior experience with driving after cannabis use changes the dynamic, and self-assessment isn't reliable enough to be the sole safeguard. Together with RTHC-00092 (Italian DUI blood draw delays) and RTHC-00124 (Wisconsin DUI delays), these studies form a driving safety cluster documenting both the problem (impaired driving) and potential solutions (self-assessment, better testing).
Replication
Not stated in abstract.
Funding
Not reported in abstract.
Conflicts of interest
Not reported in abstract.
Questions still open
- Would higher THC doses (reflecting current market potency) produce larger gaps between self-assessment and actual impairment? Could a brief validated app-based assessment supplement self-judgment? How should public health campaigns frame self-assessment — as useful but imperfect? Does tolerance from regular use improve self-assessment accuracy or simply reduce impairment at given doses?
Read the original research
Predicting changes in driving performance in individuals who use cannabis following acute use based on self-reported readiness to drive.
Accident; analysis and prevention, 195, 107376
Accident Analysis and Prevention is a reputable journal focusing on traffic safety and accident prevention research.
Citation
Miller, Ryan; Brown, Timothy; Schmitt, Rose; Gaffney, Gary; Milavetz, Gary. (2024). Predicting changes in driving performance in individuals who use cannabis following acute use based on self-reported readiness to drive.. Accident; analysis and prevention, 195, 107376. https://doi.org/10.1016/j.aap.2023.107376