A laboratory experiment from Brown University found that smoking cannabis can reduce short-term alcohol intake in a controlled setting, but researchers caution against treating cannabis as a proven substitute for alcohol in real life.
The study recreated a bar environment—complete with draft beer, spirits, ambient lighting, and a couch—to mimic typical social drinking. Researchers enrolled 138 mostly non-Hispanic white young adults in their mid-20s who regularly used both cannabis and alcohol. Each participant attended three sessions and smoked cannabis cigarettes in randomized order: a high-THC cigarette (7.2%), a low-THC cigarette (3.1%), and a very low-THC placebo (0.03%). After smoking, participants had two hours in the bar lab and could consume up to eight “mini-drinks” (enough to raise blood alcohol concentration to about 0.10 g/dL). For every mini-drink they declined, they received $3.
Compared with the placebo condition, smoking the higher-THC joint was associated with a 27% reduction in alcohol consumed; the lower-THC joint produced a 19% reduction. Participants who used the higher-THC product also delayed their first drink by roughly 11 minutes—about a 48% longer wait time. Measures of craving showed mixed outcomes: high-THC cannabis reduced immediate post-smoking cravings in some measures, but the lower dose did not, and neither dose reliably reduced cue-induced craving when participants were shown drink-related images or their preferred beverage.
These results are consistent with another recent lab study that reported roughly a 25% drop in alcohol self-administration after cannabis use. Investigators, including lead author Jane Metrik, PhD, and outside experts see the replication as encouraging but emphasize major limits. The findings describe brief effects inside a tightly controlled experimental setting and do not prove that people will drink less if they use cannabis in everyday life. Real-world behavior is shaped by social pressures, stress, marketing, and other cues that the lab cannot fully reproduce. Long-term outcomes—such as whether cannabis substitution affects overall health or increases the risk of cannabis use disorder—are not addressed by these short-term trials.
Public-health implications are complicated. Interest in approaches like “California sober,” where people replace or reduce alcohol with cannabis, has grown. Alcohol causes substantial harms: the CDC estimates about 178,000 U.S. deaths annually are tied to excessive alcohol use. Legalized cannabis is more recent, and its population-level risks and benefits remain less well understood. Some proponents see cannabis as a less harmful option; many clinicians caution against framing it as harmless.
Clinicians not connected to the study warned that lab findings could be misconstrued as endorsing one intoxicant over another. They noted that terms such as “California sober” encompass widely different behaviors—from occasional low-THC use to frequent high-potency smoking—with very different risk profiles.
In summary, this controlled experiment suggests that smoking cannabis before a drinking opportunity can reduce immediate alcohol consumption and delay the first drink in the laboratory. But researchers call for longer-term, real-world studies before recommending cannabis as a strategy to cut alcohol use, and they remind readers that cannabis carries its own risks and that effects vary by person and context.