Cannabis raises the risk of psychosis; it does not simply cause it. The clearest evidence points to a dose-response relationship, where heavier use of higher-potency cannabis carries meaningfully more risk than occasional, lower-potency use, and where most people who use cannabis, even often, never develop a psychotic disorder at all. [marconi-2016-metaanalysis] Both popular positions on this overstate something: “it’s harmless” ignores a real and replicated pattern, and “it causes psychosis” flattens a probabilistic relationship into a certainty the evidence does not support. Our psychosis pillar covers what psychosis is more broadly, including the other risk factors cannabis interacts with.
Does weed cause psychosis?
The honest answer is that cannabis use is associated with an increased risk of psychosis, and for at least some of that association the evidence is consistent with a genuine causal contribution, though it cannot fully rule out that people already vulnerable to psychosis are also more likely to become heavy users. That is a more careful claim than “causes,” and it is also a more careful claim than “no link exists.”
A 2016 meta-analysis pooling the available studies found that cannabis users overall had a higher risk of psychotic outcomes than non-users, and that risk rose further with heavier and higher-potency use, describing a dose-response gradient rather than a flat, all-or-nothing effect. [marconi-2016-metaanalysis] A dose-response pattern, where more exposure tracks with more risk, is one of the classic signs researchers look for when weighing whether an association might be causal, because it is harder to explain by a simple shared-cause coincidence than an association that does not scale with exposure.
None of this means cannabis is a sufficient cause on its own. Most people who use cannabis, including most who use it heavily, do not develop psychosis. What the evidence supports is that cannabis is one contributing factor among several, and that its contribution is larger for some patterns of use and some people than others.
How much cannabis is actually risky?
Frequency and potency both matter, and the risk is concentrated at the heavy, high-potency end of use rather than being spread evenly across anyone who has ever tried it. A large multi-country case-control study across eleven sites in Europe and Brazil found that people who used cannabis daily had roughly three times the odds of a first psychotic episode compared with people who had never used cannabis, and that daily users of high-potency cannabis specifically had roughly four times the odds, a substantially larger association than daily use of lower-potency cannabis showed on its own. [diforti-2019-europe] Occasional use showed a far weaker association in the same study.
These numbers are an illustration of the general dose-response shape described by Marconi et al. (2016) and Di Forti et al. (2019), not the exact figures either study reported. The real studies found daily use of high-potency cannabis carried roughly three to four times the odds of psychosis compared with never-users; this chart is not a substitute for those reported values.
The takeaway from both studies together is not a single risky number of joints a week. It is that the combination of frequency (closer to daily than occasional) and potency (closer to high-THC concentrate or strong resin than low-strength herbal cannabis) is what the strongest associations track, more than any single measure alone.
Who is actually at risk?
Cannabis does not act in isolation, and the same pattern of use carries different risk for different people. Risk concentrates where heavy, high-potency use meets other vulnerability already present: a family history of psychosis or schizophrenia, earlier psychotic-like experiences of one’s own, or significant childhood adversity, which is itself an established, independent contributor to psychosis risk. Starting cannabis use in adolescence, while the brain is still developing, is also associated with greater risk than starting a comparable pattern of use in adulthood.
This is the sense in which cannabis functions as a risk-multiplier rather than a standalone cause. Someone with no family history, no early psychotic-like experiences and no other major vulnerability who uses cannabis occasionally is at a very different level of risk than someone carrying several of those factors who uses high-potency cannabis daily from their mid-teens. The population-level association is real and worth taking seriously; it is not evenly distributed across every individual who has ever used cannabis.
Does quitting help?
For someone who has already had a psychotic episode, continued heavy cannabis use is linked to a rockier course: more relapses, more hospital admissions, and slower or less complete recovery in some studies. Stopping or substantially cutting back is linked to a better trajectory. This is one of the genuinely modifiable factors in the whole picture, distinct from family history or age of onset, which cannot be changed after the fact, and it is a standard part of what clinicians discuss with someone in early treatment.
Quitting does not promise a particular outcome, and it does not undo an episode that has already happened. But among the handful of levers available once psychosis has started, continued heavy cannabis use is one clinicians consistently flag as worth addressing directly rather than leaving unmentioned.
Is there a way to check whether this applies to me?
No screener on this site, or published anywhere responsibly, can diagnose psychosis or establish that cannabis specifically caused it in a particular person. That determination needs a clinical assessment, which also checks for other causes that have nothing to do with cannabis. Our self-assessment hub lists the screeners we do publish, for anxiety, depression, sleep, burnout and loneliness, and none of them covers psychosis; there is no substitute here for a clinician’s evaluation.
Worth talking to a doctor about?
This is not a diagnostic tool and cannot tell you whether cannabis is causing psychosis, in yourself or anyone else. It is a prompt for what is worth describing to a doctor, promptly, rather than waiting to see if it settles.
0 of 5 ticked
This combination, alongside heavy cannabis use, is worth a prompt professional evaluation rather than waiting. Only a clinical assessment can say what is actually happening, and earlier assessment is consistently linked to better outcomes.
Enough here to be worth a prompt conversation with a doctor, mentioning the cannabis use specifically. This is not a diagnosis of anything.
If something else feels off, it is still worth raising with a doctor on its own terms, cannabis-related or not.
No screener on this site diagnoses psychosis; only a clinical assessment can. The hub lists what we do cover.
When to seek help
Unusual beliefs that others do not share, hearing or seeing things other people say are not there, or a friend or family member undergoing a serious change in personality or behaviour alongside heavy cannabis use are all worth acting on promptly rather than waiting to see whether they pass. Speak to a doctor rather than a search engine or a forum thread. Earlier assessment and treatment for psychosis is consistently linked to better long-term outcomes, which is the practical reason not to delay.
If you or someone else is at immediate risk, contact your local emergency services or a crisis helpline.
How MyFreud can help
MyFreud cannot diagnose psychosis and will not tell you whether cannabis is the cause in a specific case; no app can. What daily tracking can do is make a pattern visible, when sleep collapsed, when use increased, when withdrawal started, which is exactly the kind of concrete timeline a clinician finds useful when someone finally does get assessed.
Download MyFreud and start today: App Store or Google Play.