Cannabis users show double the odds of committing violence

Cannabis is one of the most widely used drugs in the world, and its link to adverse mental health outcomes, including psychosis, is well documented. Psychosis in turn is frequently associated with aggression, which raises a question research had not clearly answered: does cannabis use itself carry an increased risk of violence, either as perpetrator or as victim? Pinning that down matters for public health, for psychiatric care, and for anyone weighing the risks of regular use. This review set out to synthesise everything published on the question and put a number on the association.

The researchers conducted a sweep of the literature following PRISMA guidelines, searching MEDLINE, EMBASE and PsycINFO for studies published up to November 2024. They identified 63 peer-reviewed studies covering 265,079 people and pooled the results using random-effects models. The analysis examined violence from both directions: the odds that cannabis users perpetrate violent acts, and the odds that they become victims of violence. It also split the results by population, comparing psychiatric patients with the general population, and by study design, separating studies that tracked people over time from one-off snapshots.

Cannabis users had significantly higher odds of perpetrating violence: 2.49 times among psychiatric patients and 2.05 times in the general population. The association held across designs, at 1.17 in studies that followed people over time and 2.37 in snapshot studies, a notable gap in size. The effect was especially large for violence resulting in a criminal conviction, at 3.75 times the odds in the general population and 4.21 times among psychiatric patients. Cannabis use was also associated with 1.49 times the odds of becoming a victim of violence, a risk more pronounced in females and mixed-gender cohorts than in males.

The link between cannabis and psychosis was already established, but the specific relationship with violence had remained unclear, with individual studies pointing in different directions. Pooling 63 studies makes the association hard to dismiss, though it cannot establish that cannabis causes violence, and the much smaller effect in studies that followed people over time suggests the snapshot figures may overstate the true relationship. The authors argue the findings highlight cannabis use as a potential target for violence-reduction efforts and support routine assessment of cannabis use among psychiatric patients. Why the victimisation risk was higher for females remains an open question.

If you use cannabis or are weighing whether to, this belongs in the risk picture: across a very large pooled sample, use was associated with roughly double the odds of committing violence and a meaningfully raised chance of being a victim of it, particularly for women. It does not mean any individual user will become violent, and an association of this kind cannot untangle the drug's effects from the lives and circumstances of the people who use it. It is, however, a reason to treat claims that cannabis carries no behavioural risks with some scepticism.