PTSD is a qualifying condition in most state medical cannabis programs. It is one of the most common reasons American veterans use cannabis, often with the encouragement of advocacy groups and dispensary staff who believe they are helping. It is also the condition in this entire library where the distance between how widely cannabis is used and how little controlled evidence supports it is at its widest.
In March 2026, researchers at the University of Sydney published the most comprehensive synthesis of the question yet attempted. They screened 5,774 studies and pooled 54 randomized controlled trials covering 2,477 participants across 45 years of research, testing cannabinoids as a primary treatment for mental disorders. For PTSD, they found no significant benefit. The same paper found no benefit for anxiety disorders, psychotic disorders, OCD, or anorexia, and noted that no randomized trial had ever tested cannabinoids for depression at all. The authors were careful about the limits of what they had to work with: 24 of the 54 trials carried a high risk of bias, the median trial enrolled 31.5 people, and a fifth raised conflict-of-interest concerns. That is a thin literature being summarised rather than a large clinical dataset, and the direction it points is still consistent.
The dedicated trial deserves its own paragraph, because it is widely cited as evidence in favour and it was not. In 2021, a team including Bonn-Miller and Sisley ran the first randomized placebo-controlled trial of smoked cannabis for PTSD, funded through MAPS, with 80 veterans receiving three weeks of one of three active cannabis preparations or placebo. Every group improved, including placebo. No active preparation outperformed placebo on PTSD symptom severity. The cannabis was well tolerated, which is a real finding, and the efficacy result was null. When a source cites Bonn-Miller 2021 as demonstrating cannabis works for PTSD, that source has not read past the title.
The one part of PTSD where cannabis has a thread of support
The exception is sleep, and it is narrow enough to state precisely. In 2015, Jetly and colleagues ran a small crossover trial of nabilone, a synthetic THC analogue, in ten Canadian military men with PTSD-associated nightmares. Nightmare frequency fell. The trial included no assessment of PTSD symptomatology beyond the nightmares, and ten male participants is a preliminary finding rather than a foundation. It remains one of the few positive controlled results in this field.
Systematic reviews land in the same narrow place. The reviews summarised by Utah's state medical cannabis programme concluded that cannabinoids show a signal for the sleep disturbance and intrusive symptom clusters of PTSD without improving overall symptom severity, and that the certainty of evidence for improvement in disturbing dreams with nabilone is very low. The signal exists. It is small, it is specific to nightmares and sleep, and it does not extend to the hyperarousal, avoidance, and mood symptoms that make up most of the disorder.
This matters for how to use this page. If broken sleep and nightmares are the dominant problem, the evidence and the dosing detail live on our page covering PTSD-related sleep disturbance, which treats that symptom cluster as its own clinical target. That page exists because the sleep component is the part with support behind it. The rest of PTSD is what this page covers, and the honest answer for the rest is different.
Why cannabis and trauma-focused therapy pull in opposite directions
This is the part of the page with the most practical weight, and it gets less attention than it deserves.
PTSD is maintained by avoidance. The disorder persists because the traumatic memory and its reminders are experienced as dangerous, so the person avoids them, and the avoidance prevents the nervous system from learning that the memory is a memory rather than a threat. Every effective trauma-focused therapy is built to reverse that. Prolonged exposure has people approach the memory and the avoided situations in a controlled way. Cognitive processing therapy works on the beliefs the trauma installed. EMDR processes the memory while the person holds it in mind. All three depend on the person contacting distressing material and staying with it long enough for extinction learning to happen.
Cannabis is efficient at blunting distress. That property is the reason people with PTSD reach for it, and it is the reason it sits awkwardly alongside the treatments that resolve the condition. Using cannabis to avoid contact with trauma-related distress is chemical avoidance, functionally similar to the behavioural avoidance the therapy is trying to undo. The concern is not theoretical. An individual patient meta-analysis drawing on randomized trials of evidence-based PTSD treatments, part of the Project Harmony collaboration, found poorer treatment outcomes among participants reporting cannabis use, a pattern echoed in studies of cognitive behavioural therapy for anxiety disorders more broadly.
None of this means a person using cannabis cannot benefit from therapy. It means the two interact, the interaction runs in an unhelpful direction, and the therapist needs to know. Anyone entering trauma-focused treatment should tell their clinician about cannabis use so the timing and the role it plays can be worked into the plan rather than running alongside it unexamined.
Why the trials and the lived experience disagree
Plenty of people with PTSD report that cannabis helps them, and it would be both dismissive and inaccurate to treat that as imagination. Two things are true at once, and the gap between them has a reasonable explanation.
Acute relief is real. Cannabis reduces hyperarousal, quiets intrusive thinking, and makes the evening tolerable for a few hours. That is a genuine pharmacological effect and people notice it. What randomized trials measure is different: whether symptom severity over weeks improves more with the drug than with placebo. A substance can produce reliable short-term relief without altering the trajectory of the illness, and PTSD appears to be a case where that distinction holds.
The placebo result in the smoked cannabis trial is instructive. Everyone improved, including the placebo group, by a clinically meaningful amount. Participation in a structured trial with attention, monitoring, and the expectation of help produced most of the benefit that participants attributed to cannabis. Real-world evidence tells a more favourable story, and it comes with selection built in. A study of 238 people with PTSD seeking cannabis-based medicinal products through a UK registry found improvements at three-month follow-up, in a population that chose cannabis, expected it to work, and remained in treatment. Both findings are real, and the randomized comparison is the one that isolates the drug.
There is also a safety signal worth stating plainly. The systematic reviews summarised by Utah's programme noted adverse events across the literature including worsening suicidal thinking and violent behaviour, concentrated among people who had both PTSD and cannabis use disorder. PTSD carries elevated suicide risk on its own, and heavy cannabis use in this population has been associated with worse outcomes rather than better ones. Anyone whose use has escalated, or who notices mood worsening rather than improving, should treat that as a reason to talk to a clinician.
If you are going to use it anyway
Many readers will, and a page that stops at the evidence verdict without practical guidance serves them badly.
Lean CBD, keep THC low. The CBD evidence in PTSD is thin but not empty: a controlled study by Bolsoni and colleagues found CBD reduced the anxiety and cognitive impairment triggered by recalling traumatic events in people with PTSD, which is a mechanism-relevant result even at small scale. CBD carries no intoxication, no dependence liability, and none of the biphasic anxiety risk that makes THC unpredictable in this population. A starting point is 25 to 50mg of CBD in the evening with little or no THC, understanding that the researched anxiety doses run far higher and that none of this has been tested in PTSD at scale.
Be careful with THC, and be especially careful with dose. The trauma population overlaps heavily with the panic-prone population, and THC above a modest dose provokes anxiety rather than relieving it. An edible commits you for hours with no way to stop it. If you use THC at all, stay at or below 2.5 to 5mg and pair it with more CBD.
Watch the pattern, not the dose. The signals that cannabis has become part of the problem are recognisable: needing it to get through evenings, escalating amounts, using it to avoid reminders or conversations, and irritability or anxiety in the gaps between doses. Cannabis use disorder is more common in people with PTSD than in the general population, and the withdrawal symptoms overlap with PTSD symptoms closely enough that the drug can start to look like the treatment for a problem it is sustaining.
And keep the priority straight. If you are using cannabis and not in trauma-focused therapy, the therapy is the thing with the evidence behind it. Access is often the real barrier rather than willingness, and the VA, community mental health services, and telehealth options have all expanded delivery of prolonged exposure and cognitive processing therapy in recent years. That is where the recovery data lives.