Medical cannabis collection · Sources reviewed 11 October 2026
Research on CBD in autistic people needs a more precise question than whether it “treats autism”. A study might examine anxiety, sleep, distress, disruptive behaviour or social responsiveness. These are different outcomes, and a change on one questionnaire does not establish a general improvement in wellbeing or everyday participation.
On this page
What the systematic reviews tell us
A 2025 systematic review brought together seven studies involving 494 participants. It described preliminary benefits but also substantial differences between studies, products and outcome measures. That makes a single pooled claim about CBD for autism difficult to defend. Pereira and colleagues, 2025.
A broader living review of paediatric cannabis research also shows why evidence must stay attached to its indication. Much of the formal interventional literature concerned epilepsy. Success in a defined epilepsy syndrome cannot be transferred to an autism-related difficulty simply because some children have both diagnoses. Paediatric evidence review, 2025. Its underlying search period is also important: a recent publication date does not mean it includes every recent trial.
The 150-participant trial had mixed primary outcomes
A 2021 placebo-controlled trial enrolled 150 participants aged 5–21. It tested two preparations containing CBD and THC in a 20:1 ratio, including a whole-plant extract. One co-primary outcome, the Home Situations Questionnaire, did not differ significantly between groups. On the clinician-rated global improvement measure anchored to disruptive behaviour, 49% receiving whole-plant extract were much or very much improved, compared with 21% receiving placebo. Aran and colleagues, 2021.
A secondary social-responsiveness measure also improved, while a parenting-stress measure did not. Sleepiness and decreased appetite were reported more often with the cannabinoid preparations. Describing the trial only as successful would omit an important negative primary result; describing it as entirely negative would omit the other findings.
“Evidence for efficacy of these interventions are mixed and insufficient.”
Aran and colleagues, 2021The authors’ conclusion reflects the combination of encouraging and negative findings. Both tested preparations contained THC as well as CBD.
A newer CBD trial also missed its primary outcome
A randomised crossover trial published online in December 2025, with print publication in 2026, studied CBD with terpenes in children aged 5–12. Twenty-nine completed the trial. The primary outcome, the Social Responsiveness Scale, did not show a statistically significant benefit. Some secondary measures of social relating, anxiety and parental stress improved; adaptive functioning did not. Parrella and colleagues.
These secondary findings are worth following up. They do not convert a negative primary result into confirmation of a broad treatment effect. A small study also leaves considerable uncertainty about who might benefit, how consistently and for how long.
What an open-label study can add
A 2026 six-week CBD study included 23 autistic children and adolescents with fluent language and IQ scores of at least 80. Ten met its response criterion, and several measures improved. Everyone knew treatment was being given, and there was no placebo group. Lawson and colleagues, 2026.
The study can inform feasibility, tolerability and future trial design. It cannot separate treatment effects from expectations, other changes or ordinary fluctuation. Its participants also do not represent every autistic person: findings in verbally fluent young people should not silently be generalised to people with different communication or support needs.
Choose goals that matter to the person
A reduction in visible behaviour can have several meanings. It might reflect relief from distress, but it might also reflect sleepiness or reduced participation. A useful assessment asks about comfort, communication, agency and access to valued activities, as well as the concerns of families and carers.
Where possible, the autistic person’s own account should guide what counts as improvement. For someone who cannot easily describe their experience, careful attention to their usual ways of communicating and to changes in daily function becomes especially important. A quieter presentation alone is an inadequate definition of wellbeing.
What remains unresolved?
The present literature supports continued investigation of particular difficulties. It does not establish CBD as a general treatment for autism, and it provides no basis for replacing an existing care plan with an untested retail product. Epilepsy treatment, when relevant, should remain a separate clinical decision with its own evidence and monitoring.
References and further reading
- Pereira et al. (2025). Systematic review of cannabidiol-based treatment in autism spectrum disorder.
- Living systematic review of cannabis-based medicines in children and young people (2025; underlying searches through 2023).
- Aran et al. (2021). Cannabinoid treatment for autism: a proof-of-concept randomized trial.
- Parrella et al. (online 2025; print 2026). Randomised placebo-controlled crossover trial of CBD with terpenes in autistic children.
- Lawson et al. (2026). Open-label cannabidiol study in autistic children and adolescents.
Educational information, not an individual prescribing plan. Decisions about medicines belong with the treating clinician. External reference links open in a new tab.