Cannabis the Morning After: Why a Null Group Comparison Does Not Clear Every Driver
Testing frequent cannabis users the morning after smoking did not reveal worse average performance than controls on two cognitive tasks. Within the cannabis group, however, higher THC concentrations and higher-potency products were associated with poorer performance on selected measures. Those findings can coexist. Neither supports automatically labeling every driver impaired—or clearing every driver—based only on elapsed time.
Matheson and colleagues studied 65 adults who used cannabis frequently and 65 age- and sex-matched controls. Their original 2026 observational study appears in Drug and Alcohol Dependence Reports, doi:10.1016/j.dadr.2026.100432.
What did the morning-after assessment involve?
Cannabis participants reported smoking four to seven days per week. They selected legally purchased pre-rolled cannabis, smoked at home, and attended testing approximately 12–15 hours later. Controls reported no cannabis use in the preceding month. The investigators assessed verbal recall, Trail Making performance, subjective effects, and cannabinoids in whole blood and oral fluid.
This was human research with naturalistic exposure, not randomized assignment to a fixed THC dose or a blinded placebo condition. The cannabis products averaged 30% THC, and some were infused with concentrates. Product choice, smoking behavior, and participant characteristics could influence the results.
The significant and nonsignificant results answer different questions
After adjustment for age, sex, and education, the cannabis group did not differ significantly from controls on the cognitive outcomes. That is a comparison of group averages on the measured tasks.
Among cannabis participants, higher whole-blood THC was associated with poorer delayed verbal recall and slower performance on parts of the Trail Making task. For example, the correlation between whole-blood THC and delayed recall was −0.40. This describes an association across people within that sample; it does not establish a diagnostic threshold for an individual.
Participants who smoked infused products also performed worse on delayed recall and word retention than those who smoked noninfused products. Other recall measures and Trail Making comparisons between those product subgroups were nonsignificant. A balanced interpretation should retain those negative findings rather than extending the observed memory differences to all cognitive functions.
Feeling an effect was a separate outcome
The cannabis group reported higher ratings on several visual-analogue questions, including drug effect and drowsiness. Most mood and drug-inventory subscales did not differ significantly. Subjective experience, performance on a particular task, and a measured cannabinoid concentration therefore should remain separately identified.
It would overstate the findings to claim that the participants had no residual effects simply because the cognitive group comparison was negative. It would also overstate them to turn a subjective rating into a demonstrated failure of actual driving.
Whole blood and oral fluid are not interchangeable
The study measured cannabinoids in whole blood, not separated plasma or serum. Oral-fluid THC showed associations with certain Trail Making outcomes, but not the same pattern across all memory measures. Neither matrix supplied a validated individual impairment classification.
The inactive carboxy metabolite, THC-COOH, must also remain distinct from parent THC. A correlation involving a metabolite does not establish that the metabolite itself caused the performance difference. For the broader evidentiary framework, see our article on cannabis exposure, impairment, and crash risk.
Why this study does not establish a waiting period
The assessment covered a limited set of tasks and a selected population of healthy, frequent users. It does not automatically generalize to occasional users, medical patients, edibles, other drugs taken with cannabis, or every driving demand. The observational design also cannot fully separate residual drug effects from pre-existing group differences or effects associated with habitual use.
For a case-specific opinion, the useful questions concern the actual product and route, timing, other substances, specimen matrix, and documented performance. A group-level null result is relevant evidence with limits, not a universal clearance after a specified number of hours.
Primary research
Matheson J, Zakala C, Zhao S, et al. Cognitive performance and subjective effects the morning after last use of smoked cannabis by adults who use cannabis frequently: An observational study. Drug Alcohol Depend Rep. 2026;19:100432. https://doi.org/10.1016/j.dadr.2026.100432.
