Last updated 16 August 2026 · Originally published 19 July 2022
Of everything CBD is sold for, epilepsy is where the evidence is strongest by some distance. Randomised trials, regulatory approval, years of real-world use, and children whose seizures are better controlled than they were.
Cannabidiol can reduce seizures in several severe, drug-resistant epilepsies. It does so at doses far higher than most people take, and the difference between a prescription and a shop-bought bottle is dose, consistency and supervision rather than the molecule itself.
How CBD became a licensed medicine
The order of events is worth knowing, because it’s the reverse of how these things usually go.
Small controlled trials of CBD in epilepsy date back to 1980, but they were tiny and inconclusive.[1] What changed things was families giving their children artisanal CBD oil for Dravet and Lennox-Gastaut before the modern, adequately powered trials existed. The results they reported — and the publicity around cases like Charlotte Figi’s — accelerated public interest and regulatory pressure enormously.
What emerged was Epidiolex: purified CBD, made to pharmaceutical standards, prescribed and monitored. In the United States it’s approved for seizures in three conditions — Lennox-Gastaut syndrome, Dravet syndrome and tuberous sclerosis complex — in patients from one year old. European indications and age limits differ.
All three are rare, severe, and start in childhood. They’re also famously difficult to treat — families often reach CBD after several other drugs have failed.
In the trials, it worked — though the results vary by condition. In Lennox-Gastaut syndrome, drop seizures fell by roughly 37–44% with CBD against around 17–22% on placebo.[2] In the first Dravet trial, convulsive seizures fell by 39% against 13%.[3] In tuberous sclerosis, the reduction was about 49% against 27%.[4]
It isn’t only those three conditions
The approved indications are narrower than the actual use, and that’s worth knowing.
Neurologists also prescribe CBD off-label for other drug-resistant epilepsies, and a growing body of real-world evidence suggests some of these patients benefit. A prospective study of children starting CBD at a tertiary centre, including off-label use, found strong retention at 18 months — a pragmatic indication that patients and clinicians considered it useful and tolerable enough to continue.[5]
Analyses presented in 2025 covered patients with Aicardi syndrome, CDKL5 mutations, Dup15q, myoclonic-atonic epilepsy and other rare syndromes, and reported meaningful seizure reductions. A preliminary conference analysis of 50 patients, mostly with focal epilepsy, found nearly one in five achieving at least six months of seizure freedom. Across the cohort, however, 52% eventually discontinued CBD, and the work was funded by the manufacturer.[6]
In one long-term retention review, three-quarters of patients were still taking CBD at final assessment, and more than half were able to stop or reduce another anti-seizure drug.[7]
This evidence is encouraging but less certain than the randomised data behind the three approved conditions. The label marks where CBD has been tested most rigorously — not necessarily every epilepsy in which it may prove useful.
Why the trial dose matters
Here’s the practical gap, and it isn’t about the molecule.
The dose. Prescription CBD is dosed by body weight. For Lennox-Gastaut and Dravet it’s generally maintained at 10–20 mg per kilogram a day; for tuberous sclerosis the recommended maintenance dose is 25 mg/kg.[8] For a 30 kg child that’s 300–750 mg daily depending on the condition. A retail oil might offer 25 mg in a serving.
That’s not a small dose of an effective medicine. It’s a fraction of one. Anyone giving a child a couple of dropperfuls and concluding CBD doesn’t work for seizures has not tested the question.
Consistency. Seizure control depends on a steady blood level. A product whose content varies between batches is a real problem here in a way it isn’t for someone taking CBD for stiff knees. This is the strongest argument for pharmaceutical-grade material — not that the molecule is better, but that you know exactly how much of it you’re giving.
Monitoring. Patients on prescription CBD have their liver function checked, because CBD at these doses can raise liver enzymes. At these doses that’s a sensible precaution whatever the source.
The interaction that matters most is with clobazam, a common anti-seizure drug. CBD raises levels of its active metabolite, which is why drowsiness is far more common in patients taking both — and why doctors often reduce the clobazam dose when starting CBD. That’s a managed clinical decision, and it applies equally to any CBD product.
What about full-spectrum oil?
The obvious question, and the honest answer is that no randomised trial has compared purified CBD with a standardised full-spectrum oil in epilepsy.
What exists is observational. Some reports have found patients doing well on lower reported CBD doses of whole-plant extract than of isolate — which is what you’d expect if the other cannabinoids and terpenes contribute something. Differences between products, patients and methods mean those reports can’t establish an entourage effect, but they’re consistent enough to be interesting, and it’s worth remembering that the artisanal preparations which prompted the modern trials were full-spectrum, not isolate.
The molecule is the same. The products aren’t necessarily interchangeable: other constituents may affect absorption and effects, a retail lab report doesn’t provide the manufacturing controls or stability data a licensed medicine has, and at epilepsy-level CBD doses the accompanying THC stops being trivial.
None of that shows full-spectrum oil doesn’t work. It means nobody has established whether it works as reliably, or whether an apparently equivalent CBD dose produces an equivalent result. Given how much is riding on consistency here, that gap matters more in epilepsy than anywhere else on this site.
Where this leaves it
Cannabidiol can reduce seizures in several severe, drug-resistant epilepsies. That’s as well established as anything in this field, and it reached modern medicine largely because families tried it before the large, decisive trials had been run.
What the pharmaceutical version adds is a guaranteed dose, a known purity and a clinician watching. Those are worth having in a condition like this — but they’re features of the product, not of the molecule.
The lesson for everything else on this site isn’t that prescription CBD contains a different kind of cannabidiol. It’s that the evidence belongs to a particular formulation, at a weight-based dose of roughly 10–25 mg/kg a day, in a supervised setting. It can’t automatically be transferred to a retail oil whose contents, absorption and consistency may differ.
None of this is medical advice. Epilepsy needs specialist care, and nothing here is a substitute for it.
For more, see our guide to CBD and what the research actually shows, our guide to CBD and drug interactions, or browse Health Conditions.
References
- Cunha JM, Carlini EA, Pereira AE, et al. Chronic administration of cannabidiol to healthy volunteers and epileptic patients. Pharmacology. 1980;21(3):175–185.
- Devinsky O, Patel AD, Cross JH, et al. Effect of cannabidiol on drop seizures in the Lennox-Gastaut syndrome. New England Journal of Medicine. 2018;378(20):1888–1897. doi:10.1056/NEJMoa1714631
- Devinsky O, Cross JH, Laux L, et al. Trial of cannabidiol for drug-resistant seizures in the Dravet syndrome. New England Journal of Medicine. 2017;376(21):2011–2020. doi:10.1056/NEJMoa1611618
- Thiele EA, Bebin EM, Bhathal H, et al. Add-on cannabidiol treatment for drug-resistant seizures in tuberous sclerosis complex: a randomized clinical trial. JAMA Neurology. 2021;78(3):285–292.
- Assessing real-world efficacy, safety and 18-month retention rates of cannabidiol in individuals with drug-resistant epilepsies. Epilepsia Open. 2025.
- Long-term use of cannabidiol (Epidiolex) in patients with refractory epilepsy: the UM experience. Preliminary analysis presented at the American Epilepsy Society annual meeting, December 2025. Funded by Jazz Pharmaceuticals.
- Gaston TE, et al. Real-world, long-term evaluation of the tolerability and therapy retention of Epidiolex (cannabidiol) in patients with refractory epilepsy. Epilepsy & Behavior. 2023;141:109141. doi:10.1016/j.yebeh.2023.109141
- Epidiolex (cannabidiol) oral solution: US prescribing information, revised May 2026. Jazz Pharmaceuticals.
About this article
Substantially rewritten in August 2026, consolidating our earlier reporting on CBD and seizures. This update sets out what the trials showed, how CBD came to be licensed, and why dose and consistency matter more than which product the cannabidiol came from. With earlier reporting by Nick Congleton and Lydia Kariuki.
