Does Cannabis Help PTSD? What the Research Actually Shows

Patients ask inside the first ten minutes, usually sideways. Does this stuff work, or am I about to waste my money.

The honest answer takes longer than anyone wants it to.

It starts with an accident. A psychiatrist in Nova Scotia named George Fraser was prescribing nabilone, a synthetic cannabinoid that comes as a capsule and is not something you would find in a dispensary, for a condition unrelated to any of this, and he noticed his patients were reporting that their nightmares had stopped.

He wrote up 47 cases in 2009. No control group, nobody blinded, one clinician describing what he had seen in his own practice over several years.

Nightmares suppressed in roughly seven of ten.

That is how a fair amount of medicine starts. It is also not evidence, and Fraser did not claim it was.

The trial everyone cites

The controlled version came six years later. Ten Canadian soldiers whose nightmares had outlasted everything else they had been given, randomized to nabilone or to placebo, seven weeks on each with a two-week washout in between, then crossed over to the other arm. Half a milligram to start, titrated up to three.

On nabilone the distressing dreams came down in frequency and in intensity, and the result held up statistically despite the sample being what it was.

Then you read the rest of it.

The dream measure moved. The broader sleep measures did not. What the trial showed was a change in one symptom, and the night around that symptom stayed roughly where it had been.

Two things about that study we say out loud, because almost nobody does. Nabilone is a prescription capsule. And ten men is ten men, which we had wrong in our own heads for a while, telling people it was closer to thirty until we went back and looked.

The authors called it preliminary and said it would need repeating in a larger group. A multicentre trial is finally running now, using a different cannabinoid. It has not reported.

What the field says now

In March 2026 The Lancet Psychiatry published the largest analysis anyone has done on cannabinoids in mental health, pulling together 54 randomized trials across 45 years, and found no reliable evidence that cannabis treats PTSD, anxiety, or depression.

A JAMA review the same month arrived at the same place by a different route.

It did give a little ground on sleep, and it is worth being exact about how little, because this is the part that gets quoted badly. Across four insomnia trials there was no effect on insomnia symptoms and no effect on sleep quality. What moved was total sleep time.

That was the only moderate-certainty finding in the entire paper, and it stopped being significant once the weakest studies were taken out.

One more thing in there got almost no coverage. Most of the trials the review examined used licensed pharmaceutical cannabinoids at known doses. Sativex, mostly. Purified compounds.

Almost none of them tested the high-THC flower and concentrate that make up the actual market, which means the evidence base, thin as it already is, is not really evidence about what is sitting on the shelf in Ohio.

So what we take from all of it is narrower than either side would like.

Cannabis will not resolve the disorder. In some patients it does something specific to one symptom, and one symptom is frequently enough to change a week.

That last part is not from any trial. It is what we watch happen in clinic, which is a weaker kind of evidence and we will say so. Get a man three consecutive nights and the irritability comes down, concentration comes partway back, the fuse with his kids gets longer. Nobody would call that cured. He got his week back.

The part that points away from us

Cognitive processing therapy, prolonged exposure, and EMDR carry the strongest evidence in PTSD by a wide margin. More than half the people who finish a full course no longer meet criteria afterward. Nothing in a dispensary approaches that number, and we tell every patient so, usually in the first visit.

Where cannabis sometimes earns its place is in getting somebody through that door. Hyperarousal runs high enough in some people that sitting in a room and narrating the worst day of their life is not physically available to them. When something takes enough edge off that they finish the course, that has done real work.

We have watched that happen more than once. We have also watched it not happen.

So, does it work. For the disorder, no. For one symptom, in some people, sometimes, on evidence thinner than we would like.

That is what we say at the first visit rather than the fourth. What you do with it depends on which version of PTSD you actually have, which is a different conversation and one we have written about here.

---

Schedule your evaluation, in person or virtually, here. This is how We Do Cannabis Differently.

If you are in crisis or having thoughts of harming yourself, call or text 988 to reach the Suicide and Crisis Lifeline, available 24/7. Veterans can press 1.

Nabilone is a prescription medication and is not available through Ohio dispensaries. Discuss any change to your treatment with your physician.