A Next-Morning Cannabis Driving Study Found No Significant Group Deficit After Correction
Frequent cannabis users tested approximately 12–15 hours after smoking did not show a significant simulator-driving deficit relative to matched controls after correction for multiple comparisons. This observational finding does not establish a universal waiting period or prove that every participant was unimpaired.
A next-morning group comparison
The study compared 65 frequent cannabis users with 65 controls who had not used cannabis during the preceding month. Groups were matched for age and sex and were similar on several driving and sleep measures, but differed in education and racial composition.
Cannabis users smoked a preferred commercial pre-roll at home the prior evening. Timing and amount were reported or estimated rather than continuously observed. This was not a placebo-controlled dosing experiment.
Correction changed the interpretation of lane-position findings
Lane-position variability was slightly lower in the cannabis group during standard and distracted driving. Those differences did not remain statistically significant after correction for multiple comparisons. Other assessed driving measures also did not significantly differ between groups.
It would therefore be misleading to call the small lane-position differences evidence that cannabis improved driving. Conversely, a nonsignificant comparison does not demonstrate exact equivalence or exclude smaller effects that this design could miss.
Chemical and subjective findings did not supply a driving threshold
Mean reported blood THC was 2.8 ng/mL, while oral-fluid THC was measured separately. Neither blood nor oral-fluid THC significantly correlated with the driving outcomes after correction. Some associations involving 11-hydroxy-THC also failed to remain significant after correction.
Cannabis users reported higher subjective cannabis-effect ratings. Those positive subjective findings coexist with null adjusted driving comparisons; neither endpoint can substitute for the other.
What the design leaves unresolved
There was no individual drug-free baseline driving assessment against which each user’s next-morning performance could be compared. Matching and adjustment reduced some differences between groups but could not remove all pre-existing differences.
The findings concern frequent users, smoked products and the tested simulator tasks. They do not directly establish outcomes for infrequent users, edibles or every real-road situation. Collision counts were too small for analysis. Interpretation should preserve the study’s useful null result without turning it into a guarantee of safety or a precise individual recovery clock.
Related: Interpreting THC concentration and impairment.
Primary research
Driving by frequent cannabis users ‘the morning after’ last use of smoked cannabis: an observational driving simulator study. DOI: 10.1186/s42238-026-00416-w.
