Cannabis and epilepsy research points to one specific compound rather than the plant as a whole: cannabidiol, usually shortened to CBD. Randomised placebo controlled trials have shown that pharmaceutical grade cannabidiol reduces seizure frequency in two rare childhood onset epilepsies, Dravet syndrome and Lennox-Gastaut syndrome, when added to standard antiseizure medicines. In South Africa that is prescription territory, led by a neurologist, and it is not a cure for epilepsy.
Last updated:
Key takeaways
- The credible evidence on cannabis and epilepsy is about cannabidiol specifically, not about smoking or eating cannabis.
- Two large trials found real but partial benefit: fewer seizures than placebo, not seizure freedom.
- Benefit was shown as an add on to existing antiseizure medicines, never as a replacement for them.
- Side effects were common, and raised liver enzymes mean blood monitoring is part of the picture.
- In South Africa, access runs through a doctor, the scheduling rules, and where needed a SAHPRA Section 21 authorisation.
What the evidence on cannabis and epilepsy actually shows
Two double blind, placebo controlled trials published in the New England Journal of Medicine did most of the heavy lifting here, and both tested a purified cannabidiol oral solution rather than a plant product.
In the Dravet syndrome trial, 120 children and young adults with drug resistant seizures were randomly assigned to cannabidiol or placebo alongside their usual treatment. The median number of convulsive seizures per month fell from 12.4 to 5.9 in the cannabidiol group, compared with a fall from 14.9 to 14.1 on placebo. Put plainly, roughly a halving of seizures against almost no change. Seizure freedom was rare: 5 percent of the cannabidiol group became seizure free, versus none on placebo.
In the Lennox-Gastaut trial, 225 patients aged 2 to 55 were assigned to one of two cannabidiol doses or placebo. Drop seizures fell by a median of 41.9 percent on the higher dose and 37.2 percent on the lower dose, against 17.2 percent on placebo. Both differences were statistically significant.
Those two results are the reason cannabis and epilepsy is taken seriously by neurologists at all. They are also the reason the conversation stays narrow. The trials studied specific syndromes, a specific purified compound, and a specific role as an add on therapy.
| Situation | What the trial evidence supports | Strength |
|---|---|---|
| Dravet syndrome | Purified cannabidiol added to existing medicines reduced convulsive seizure frequency versus placebo | Randomised controlled trial |
| Lennox-Gastaut syndrome | Purified cannabidiol added to existing medicines reduced drop seizure frequency versus placebo | Randomised controlled trial |
| Other drug resistant epilepsies | Observational reports only, with inconsistent quality and no placebo comparison | Weak, still being studied |
| Smoked or edible cannabis | No controlled trial evidence for seizure control, and dose and purity are unknown | Not supported |
| Replacing antiseizure medicines | Never tested that way. Both trials kept patients on their usual regimen | Not supported |
Why cannabidiol and not cannabis in general
Cannabis contains hundreds of compounds. The two most discussed are cannabidiol and tetrahydrocannabinol, or THC. Cannabidiol is not intoxicating at the doses studied. THC is intoxicating, and in epilepsy it carries a particular concern, because sedation, cognitive change and anxiety are all unhelpful in someone already managing seizures.
That distinction matters practically. A cannabidiol product made to pharmaceutical standards has a known concentration, a known purity and a batch that can be traced. An unregulated oil bought informally has none of those things, which makes both benefit and harm impossible to attribute. For a condition where a missed or mis dosed medicine can trigger a seizure, that uncertainty is not a small problem. A condition with a very different evidence picture is covered in our guide to cannabis for migraines, where the 2022 reviews report fewer migraine days for some patients.
Before you start
Epilepsy is a condition where changing treatment without supervision can be dangerous. Never stop, reduce or substitute an antiseizure medicine in order to try a cannabis based product. Doing so risks breakthrough seizures and status epilepticus.
In South Africa, cannabis based medicines containing THC are prescription only through a registered doctor, usually via a SAHPRA Section 21 authorisation where the medicine is not registered locally. This page is general information for adults, including parents and caregivers, and is not medical advice. Speak to a neurologist or pharmacist before starting, stopping or changing any treatment, and about dosing.
How cannabis and epilepsy care is accessed legally in South Africa
South African rules are their own thing, and overseas coverage is a poor guide. Three separate pieces decide what is possible.
Scheduling. Under the notice published in Government Gazette 43347 on 22 May 2020, cannabidiol sits in Schedule 4, which makes it prescription only, with two narrow exceptions that fall to Schedule 0. Those exceptions are complementary medicines containing no more than 600 mg of CBD per sales pack with a maximum daily dose of 20 mg and only a low risk claim, and ingested processed products containing 0,0075 percent or less CBD. THC sits in Schedule 6, with its own narrow exceptions. A general shop bought CBD oil sold under the Schedule 0 exception is therefore capped at a daily dose far below anything used in the epilepsy trials.
Section 21. Where a cannabis based medicine is not registered in South Africa, a doctor can apply to the South African Health Products Regulatory Authority for authorisation to access it for a named patient under Section 21 of the Medicines and Related Substances Act 101 of 1965. That application is made by the prescriber, not the patient, and it is decided case by case.
The Cannabis for Private Purposes Act. Act 7 of 2024 deals with private cultivation and possession by adults. It is not a medical supply route, it does not create a legal market for treating a medical condition, and it does not apply to children. Conflating the private use law with medical access is the single most common mistake in this area.
The practical consequence for cannabis and epilepsy is straightforward. A serious attempt runs through a neurologist, a specific product, a prescription, and where relevant a Section 21 authorisation. A pharmacy dispenses it. Nothing about the private use law shortcuts that.
The risks, side effects and interactions worth knowing
The trials were candid about harms, and the pattern was consistent across both.
- Diarrhoea, vomiting, fatigue, fever and drowsiness were all more common on cannabidiol than on placebo in the Dravet trial.
- Somnolence, decreased appetite and diarrhoea were the most common adverse events in the Lennox-Gastaut trial, and were worse at the higher dose.
- Raised liver enzymes appeared in both. In the Lennox-Gastaut trial, 14 patients on cannabidiol, about 9 percent, had elevated liver aminotransferase concentrations.
- More people withdrew from the cannabidiol arm than the placebo arm in the Dravet trial.
Drug interactions deserve separate attention. Cannabidiol is metabolised by, and inhibits, several cytochrome P450 liver enzymes, which is the same machinery that clears many antiseizure medicines. The result can be higher blood levels of a co prescribed drug, producing side effects that look like the cannabidiol but are actually the other medicine. This is why baseline and follow up blood tests, and a full medicine list shared with every prescriber, are routine rather than optional.
Questions worth taking to a neurologist
A specialist appointment goes further when the questions are specific. Useful ones include:
- What is my seizure type and syndrome, and does it resemble the ones where cannabidiol has trial evidence?
- Which standard treatments have genuinely been tried, at what doses, and for how long?
- If cannabidiol were added, which of my current medicines would need level monitoring or a dose change?
- What baseline blood tests would I need, and how often would they be repeated?
- Would the product be a registered medicine, or would this require a Section 21 application, and what does that involve?
- What result, measured how, would count as success, and at what point would we stop?
Bringing a seizure diary makes that conversation considerably more productive. If you want help framing the discussion, our guide on how to talk to your doctor about cannabis covers the practical side, and medical cannabis dosage in South Africa explains why dosing is a prescriber decision rather than a look up table.
How Professor Green fits into the process
Professor Green is a South African medical cannabis platform. Registering creates an account and opens the member area, which is where the store journey begins. That is what registration does, and it is worth being precise about what it does not do.
Professor Green does not diagnose epilepsy, does not classify seizures, does not write prescriptions and does not lodge Section 21 applications. Those are acts performed by registered practitioners: a neurologist assesses and prescribes, a pharmacy dispenses, and SAHPRA decides a Section 21 application. Registration is a first step into the platform, not a guarantee of eligibility, supply or any particular product, and it is for adults.
For readers who arrived here from the legal side rather than the clinical side, our explainers on whether cannabis is legal in South Africa and the Cannabis for Private Purposes Act explained set out the current position. If you want to explore the platform itself, you can register for store access.
Frequently asked questions about cannabis and epilepsy
Can cannabis cure epilepsy?
No. Nothing in the cannabis and epilepsy evidence base shows a cure. The trials measured reductions in seizure frequency compared with placebo, while patients stayed on their existing antiseizure medicines. Seizure freedom was uncommon, and epilepsy remained a condition needing ongoing management.
Is CBD from a health shop the same as what was tested?
No. The trials used a purified pharmaceutical grade cannabidiol solution at a controlled dose. A Schedule 0 retail CBD product in South Africa is limited to a maximum daily dose of 20 mg and a low risk claim, which is a different product used in a different way.
Is cannabidiol safe for children with epilepsy?
The trials specifically included children with Dravet and Lennox-Gastaut syndromes, under specialist supervision and with monitoring. Side effects were common and liver enzymes rose in some participants. Any use in a child is a paediatric neurology decision, not a parent led one, and THC containing products are generally avoided.
Can cannabidiol replace my current epilepsy medication?
No, and it has never been tested that way. In both trials cannabidiol was added to standard treatment. Reducing or stopping an antiseizure medicine without specialist supervision risks breakthrough seizures and status epilepticus.
Does cannabidiol interact with other epilepsy medicines?
Yes. Cannabidiol affects cytochrome P450 liver enzymes that also process many antiseizure drugs, which can raise the levels of those drugs. Prescribers manage this with blood monitoring and dose adjustment, so every medicine and supplement you take should be disclosed.
What does a Section 21 authorisation involve?
Where a medicine is not registered in South Africa, the treating doctor applies to SAHPRA for authorisation to use it for a named patient. The prescriber submits the application with clinical motivation, and SAHPRA decides. Patients cannot apply for themselves.
Does the Cannabis for Private Purposes Act let me grow cannabis to treat seizures?
No. Act 7 of 2024 addresses private cultivation and possession by adults. It creates no medical supply route, sets no medical standard for a plant grown at home, and does not apply to children. Medical access remains a prescriber and pharmacy pathway.
The bottom line on cannabis and epilepsy
The honest summary of cannabis and epilepsy is narrower than the headlines and more useful than the hype. One purified compound, tested in two rare syndromes, produced a genuine reduction in seizures as an add on therapy, with real side effects and real drug interactions. Everything beyond that is either still being researched or is not supported at all. In South Africa the route runs through a neurologist, the scheduling rules, and where the medicine is unregistered a SAHPRA Section 21 authorisation. Start with the specialist appointment, and treat any product that promises more than the evidence with suspicion.
Sources and references
- Devinsky O et al. Trial of Cannabidiol for Drug-Resistant Seizures in the Dravet Syndrome. New England Journal of Medicine, 25 May 2017
- Devinsky O et al. Effect of Cannabidiol on Drop Seizures in the Lennox-Gastaut Syndrome. New England Journal of Medicine, 17 May 2018
- SAHPRA. Communication to industry, scheduling status of THC and CBD (General Notice 586, Government Gazette 43347, 22 May 2020)
- SAHPRA. Section 21 access to unregistered medicines
- Medicines and Related Substances Act 101 of 1965
- Cannabis for Private Purposes Act 7 of 2024, Government Gazette 50744, 3 June 2024
- Drug interactions of tetrahydrocannabinol and cannabidiol in cannabinoid drugs, recommendations for clinical practice
- An update on safety and side effects of cannabidiol, a review of clinical data and relevant animal studies
- World Health Organization. Epilepsy fact sheet
- Epilepsy South Africa
- Health Professions Council of South Africa, for confirming a practitioner is registered
Explore Professor Green store access
If you have been researching cannabis and epilepsy and want to look at the Professor Green platform, the next step is to create an account. Registration opens the member area. Diagnosis, prescribing and dispensing stay with your neurologist and pharmacy.
Register for Store Access

