Snake Oil Cannabis? The Scientific Evidence for Medical Marijuana & CBD


What medical conditions are marijuana / cannabis and CBD (Cannabidiol) actually effective for? A round-up of the latest studies and evidence.

Use our interactive to filter & explore different medical conditions or the particular effects of CBD oil, the other major ingredient of cannabis after THC.

» See the data for more information, links & abstracts.

Part of our deeply evidence-based Snake Oil series, currently covering supplements and superfoods.

Created with our data-visualisation software VizSweet

Our Methodology & Thinking

Our evaluation is based on the latest research and studies. We score only for human trials, preferably placebo-controlled (though it’s difficult to control for cannabis due to its distinct psychoactive effects).

Many of the studies and medical claims about cannabis today are based on cell or animal studies – the effects of which rarely translate to humans.

There’s also a lot of anecdotal evidence and single case studies out there. Understandable, since people have been self-medicating with illegal marijuana for decades. But, while these stories are heartening and promising, they don’t really count as solid, scientific evidence.

The low number of human studies is due, in part, to cannabis remaining a schedule 1 drug under US federal law. Despite legalisation in several US states, much red tape and legal obstacles block those wishing to study marijuana’s medicinal effects on humans.

Not All Studies are Equal

When evaluating media reports on medical marijuana, it’s helpful to understand the different kinds of studies.

case study – WEAK – an examination of a single case of a single patient.

pilot study – WEAK but INTERESTING – small preliminary trial to test the feasibility of a treatment or justification for a larger study.

review article – INTERESTING – summarised  results of existing trials and studies without any further analysis.

case control study – GOOD – specially targets and examines side-effects among populations already suffering with an ailment.

observational / epidemiological / population study – GOOD – statistical analysis of a large populations for patterns & effects over time, but hidden factors may influence perceived results.

controlled / clinical trial – GOLD STANDARD – randomly selected groups of people given either an active substance or a non-active ‘placebo’. Ideally ‘blind’ so neither the subjects nor the experimenters know who’s receiving placebo.

meta-analysis / meta study – VERY GOOD – critically examines & re-evaluates all existing trials & studies and mashes the results to look for consistencies & patterns.
(Cochrane is the gold standard)


This is an ongoing, evolving graphic. So let us know if we’ve missed any studies or conditions. Please cite your sources.