Along with chronic pain, mental health conditions are some of the top reasons people use marijuana for medical purposes.
But a sweeping review of cannabis studies over the past 45 years concludes there is little to no high-quality evidence showing this is effective.
The findings, published in the medical journal The Lancet Psychiatry, underscore the extent to which the public's embrace of cannabis has outpaced the scientific research.
The new analysis represents the largest effort yet to systematically parse all the data from randomized-controlled trials on cannabis and mental health. A team of researchers in Australia looked at more than 50 clinical trials, considering a wide range of conditions, formulations and types of cannabinoids.
The analysis turned up no evidence that cannabis can help with symptoms of anxiety, post-traumatic stress disorder or depression — the psychiatric conditions that medical marijuana users most often mention when asked why they're taking the drug.
Insomnia, autism and tic or Tourette's syndrome had more supporting data, though even that evidence was deemed "low quality" by the authors.
"We clearly need to do more research on cannabis medicines," says Jack Wilson, a postdoctoral research fellow at the Matilda Centre for Research in Mental Health and Substance Use at the University of Sydney who led the review. "In the absence of evidence at this time, the routine use of medical cannabis products really should be rarely justified for the treatment of mental health disorders," he told NPR.
The findings are not altogether surprising for cannabis researchers who know first-hand how challenging it is to conduct and fund well-controlled trials. For more than 50 years, cannabis has been listed as Schedule 1 drug by the Drug Enforcement Administration, though President Trump recently signaled he'd like the federal government to loosen that designation.
Even as many states moved to legalize medical and recreational marijuana, that didn't result in large investment in the kind of high-quality studies that medicine leans on when evaluating treatments.
Because the Lancet Psychiatry paper had strict criteria for which studies could be considered, the final analysis only included data from close to 2,500 patients. And for some conditions like depression, there was not a single trial available.
"It's embarrassing how little we have done in terms of data collection, given how widely this is available as a therapeutic," says Ryan Vandrey, a professor of psychiatry and behavioral sciences at Johns Hopkins University School of Medicine who studies cannabis.
Though done well, a review of this nature invariably comes with major limitations, he says. It collapses data from different products, doses, routes of administration, patient populations, and so on — and excludes the findings from long-term, observational studies and other sources of evidence.
"So it can be challenging to draw firm conclusions, especially when there's not that many actual studies or patients being evaluated," he says.
The Lancet Psychiatry study comes on the heels of another major review, published earlier this month in JAMA Internal Medicine.
That took a broader view — considering other types of studies, not just controlled trials — but came to a similar conclusion on the lack of evidence for treating mental health conditions.
It also warns of "substantial risks" in vulnerable groups, including adolescents and young adults, those at risk of substance use disorders, as well as people with bipolar disorder or psychotic disorders. There is a well-documented link between cannabis use at a young age and an increased risk of psychosis.
"What we wanted to do is convey that cannabis is not one thing because it's such a complex substance," says Dr. Devan Kansagara, a professor of medicine at Oregon Health and Science University and the Portland VA.
Kansagara, who runs a project funded by the Department of Veterans Affairs aimed at synthesizing the evidence on cannabis, says doctors need to speak with their patients who are using cannabis, while acknowledging there are still big gaps in our understanding.
One take-home message from their work is an emphasis on the dose — that high THC-containing products like gummies and concentrates, especially for those who have a serious mental illness, appear to carry the most risks.
Opting for a lower dose THC product is one "way to more closely balance benefits and harms," he adds.
Cannabis can have "vastly different physiological effects" depending on the compound, though some of that nuance gets lost in the Lancet Psychiatry study, says Ziva Cooper, a professor of psychiatry who directs the UCLA Center for Cannabis and Cannabinoids.
For example, she says the review highlights that "cannabinoids as a whole haven't been shown to be helpful for anxiety," but the picture changes when you look at specific compounds in the plant, such as cannabidiol, or CBD, which has shown promise.
"We should be open to integrating other types of data that aren't necessarily from placebo-controlled studies," Cooper says.
Vandrey has been involved in several research studies that have shown "significant clinical benefit to patients with anxiety and depression," he says. Neither were included in the Lancet Psychiatry analysis because they didn't meet the criteria.
"There's a subset of people with anxiety, depression, or PTSD that can realize tremendous benefit when they start using a cannabis product for that purpose," Vandrey says. "Now we also see that some patients try it and it really has no effect. And some patients try it and things get worse."
Copyright 2026 NPR