CBD, THC and Next-Day Impairment: What Negative Tests Do—and Do Not—Show

Medical cannabis collection · Sources reviewed 11 October 2026

A study that finds no measurable impairment deserves to be reported just as carefully as one that finds harm. The challenge is to say what the negative result establishes. CBD and THC, subjective sleepiness and objective performance, and an acute laboratory test and real-world driving are related questions rather than interchangeable ones.

Feeling sleepy and performing poorly are different measurements

A 2024 meta-analysis examined 16 human laboratory trials of acute CBD effects, measured within eight hours. The overall effect combining different outcomes was very small. When separated, subjective outcomes showed a small effect, largely involving sedation, while objective performance outcomes did not show a statistically significant pooled impairment effect. THC produced greater impairment than CBD in the available direct comparisons. Lo and colleagues, 2024.

“a small increase in subjective ratings of sedation”

Lo and colleagues, 2024

The review distinguishes how participants felt from how they performed on the objective tasks measured.

These findings provide some reassurance about the particular objective tests and conditions studied. They also explain why two descriptions—no detected performance deficit and greater reported sleepiness—can both be accurate. Neither measurement should be silently substituted for the other.

What does “next day” mean in research?

A 2023 systematic review defined next-day testing as more than eight hours after THC exposure. Across 20 studies involving 458 participants, most tests did not detect an effect. Evidence of impairment was limited and came from older studies with methodological weaknesses. Many other results were unclear, and some studies had not demonstrated sensitivity to acute impairment in the first place. McCartney, Suraev and McGregor, 2023.

The review therefore challenges the assumption that THC necessarily impairs every measured ability the following day. Its definition is a way of sorting research observations; it is not a recommendation that everyone can drive eight hours after any cannabis product. The evidence largely concerned single exposures and particular laboratory tasks.

A small insomnia trial found mostly negative tests

A 2024 randomised crossover pilot studied 20 adults with insomnia who infrequently used cannabis. Participants received a single bedtime oral preparation containing 10 mg THC and 200 mg CBD, or placebo. The next-morning cognitive, psychomotor and simulated-driving assessments began at least nine hours after dosing. These were secondary outcomes of the trial. The next-day insomnia pilot.

No difference was detected on 27 of 28 performance measures. There was a small reduction in accuracy on an easier Stroop task, and subjective sedation was greater at ten hours. The harder Stroop task did not show the same difference. The doses describe the tested preparation and are not a dosing recommendation.

This is a reassuring result within a narrow experiment. It does not establish the effects of repeated treatment, every THC:CBD ratio, other medicines, or real-road driving in a much larger and more varied population.

How to interpret a negative result fairly

“No statistically significant difference” can describe a genuinely small effect, an insensitive test, a small sample or an imprecise estimate. To distinguish these possibilities, look at the confidence interval, the number of participants and whether the task could detect a meaningful impairment.

The reverse caution matters too: testing many outcomes can produce an isolated difference by chance. A single abnormal result should be examined in relation to the prespecified analysis and the overall pattern. Neither selecting only the negative tests nor selecting only the positive test gives a balanced account.

Different people, products and circumstances

A trial with infrequent users and a single bedtime dose does not describe every person receiving long-term treatment. The route, formulation, timing, accompanying medicines and reason for use may all change which study is relevant. Pain or poor sleep can also affect baseline performance, so a clinical study needs to distinguish the illness from the treatment’s additional effect.

Average laboratory performance cannot certify an individual’s ability to drive or operate equipment. The separate article on UK prescribing and safety discusses the legal and practical context. This article’s narrower point is that the research should be read without inventing a universal waiting period from a study’s assessment schedule.

The useful unanswered questions

The available negative findings have value: they limit claims that residual impairment is inevitable. Their boundaries have equal value: they prevent limited laboratory reassurance becoming an unconditional safety promise.

Educational information, not an individual prescribing plan. Decisions about medicines belong with the treating clinician. External reference links open in a new tab.