Could smoking pot lead people to drink less alcohol?

In an elaborate and provocative experiment, scientists doled out joints and free drinks to see whether this idea — sometimes described as “California sober” — survived scientific scrutiny.

The new research offers some of the strongest data yet suggesting that smoking weed does, at least in the short term, curb how much people drink.

The findings were published in the American Journal of Psychiatry on Wednesday — and are sure to raise questions about the merits of swapping one of these substances for another, especially given growing concern in the public health field about the popularity of cannabis.

Millions of Americans use weed to treat chronic pain, but there’s little high quality research on whether it works. New findings suggest it can be effective for low back pain, on par with opioids.

And the researchers are cautious about making any recommendations based on their findings at this point.

“We’re not ready to tell people seeking treatment for alcohol, go ahead and substitute cannabis, and it will work out for you,” says Jane Metrik, a professor of behavioral and social science at Brown University who led the study.

But the research does bring scientists closer to understanding the link between these two substances, at a time when many people are leaning on marijuana to cut back on drinking, without waiting for the evidence.

“This study really moves the field forward by helping to resolve one of the unresolved questions in the literature,” says Jeff Wardell, a professor of psychology at York University. “This gives us more confidence that there’s a real effect here.”

Getting high for science

In the Brown study, the team took pains to replicate the real-life circumstances of getting stoned and drinking, while still maintaining a tightly controlled study that could point toward a causal relationship.

They constructed a laboratory that resembles a bar, with comfy seats and a tap, and ensured each participant had their preferred alcoholic beverage on hand.

“We wanted to make sure that when given the opportunity, you would be really driven to drink,” says Metrik, who ended up spending substantial time shuttling between liquor stores looking for special vintages and spirits.

The experiment included three separate sessions. In one of them, participants smoked a marijuana joint with higher levels of the psychoactive compound, THC; in another they used a lower potency strain; and the last, researchers gave them a placebo with a trace amount of THC that wasn’t enough to get them high.

After toking up in a designated smoking room, each participant spent the next two hours in the “bar lab” on their own where they had the opportunity to drink up to eight mini drinks.

The people who smoked the higher potency cannabis ended up drinking 27% less alcohol, and the lower potency about 19% less, compared to the placebo. People who used cannabis also delayed their drinking.

“It is an important signal that we’re detecting,” says Metrik. “It is telling us that cannabinoids could play a potential therapeutic role in alcohol use disorder.”

Previous research has suggested that cannabis may reduce alcohol cravings and how much people drink. But the results have been largely inconclusive, in part because the data often come from observational studies, which are less reliable and can be muddied by other factors. Animal studies have also indicated potential biological mechanisms behind weed’s effect on alcohol; however, there’s a question of how much this applies to humans.

The new study also builds on what a team of researchers in Colorado reported earlier this year in a slightly different experiment.

There, participants instead picked up their marijuana from a dispensary, smoked it at home and then visited a mobile lab parked nearby, where they were offered alcoholic beverages.

The amount people ended up drinking dropped by about 25% when they were already stoned. Cravings also went down.

“These findings are all converging on a similar story,” says Hollis Karoly, an associate professor of psychiatry at the University of Colorado Anschutz who led the study. But she points out there are still big questions about how much you can extrapolate from this new evidence.

Replacing one problem with another?

One obvious limitation is that how someone behaves when every sip they are taking is under close observation may not necessarily reflect what happens in a much looser, real-world social setting.

And beyond that, Karoly says this pattern toward reduced alcohol consumption wasn’t true for everyone in their study. In a minority of people, it actually led them to drink more.

“This really highlights the fact that individual differences matter,” she says.

Another unanswered question centers on who’s being studied.

Most of the people in the Brown trial met the criteria for “cannabis use disorder,” and about 40% for alcohol use disorder. That could indicate their “drug of choice” was not necessarily alcohol in the first place, says Rajita Sinha, a professor of psychiatry at Yale University.

For people with problem drinking, cannabis might be a potential treatment, she says. This approach, on the other hand, “may also be promoting more cannabis use and that would be problematic.”

“Once you’re in the throes of daily cannabis use, it’s very hard to kick that habit,” she says.

Those involved in this work acknowledge the tension in this line of research.

Cannabis doesn’t carry the same risks of extreme harm as alcohol, which is a leading cause of preventable death in the U.S and kills more than 170,000 people a year.

But Wardell says cannabis is clearly “not a harm-free substance,” either.

While research on its effect has not kept up with its surging popularity, studies have shown marijuana can impair cognition and memory, trigger a serious gastrointestinal syndrome that causes nausea and vomiting, increase the risk of psychosis and other psychiatric illness, and impact relationships and social functioning.

And Wardell says this latest study doesn’t actually tell us whether the negative outcomes of alcohol are worse than cannabis in the long run.

“We would have to choose which one for a given individual might be less harmful and make sure that it’s not inadvertently just replacing one problem with another,” he says.

In her role as a clinical psychologist, Metrik at Brown has seen some patients with severe alcoholism who successfully stopped drinking with the help of cannabis, though some have then gone on to develop issues with that drug.

In her mind, the problem right now is that many people are relying on cannabis to treat their alcohol problems, without any guidance.

“We see this all the time, and we don’t know what to tell them,” she says. “There’s no clear messaging,”

Copyright © 2025 NPR