Medical Cannabis Use More Than Doubled in the U.S. Between 2013 and 2020

Medical cannabis use in the United States more than doubled between 2013 and 2020, according to a peer-reviewed analysis of national survey data—but the study does not show that legalization caused the increase or that cannabis was effective for the conditions reported by users.

Published in 2023 in the American Journal of Preventive Medicine, the study by researchers associated with the Department of Veterans Affairs and Yale University examined responses from 426,092 people who participated in the National Survey on Drug Use and Health. The survey began asking about medical cannabis use in 2013.

The share of U.S. residents reporting cannabis use for medical purposes rose from 1.2% in 2013–2014 to 2.5% in 2019–2020. Researchers calculated an average annual increase of 12.9% during the study period.

The survey’s measure of medical use was specific: respondents who had used cannabis during the previous year were asked whether any of that use had been recommended by a doctor. It therefore does not necessarily indicate that cannabis was formally prescribed, obtained through a state medical-cannabis program, or used for a medically established indication.

Living in a state that had legalized medical cannabis was strongly associated with reported medical use. After adjusting for other factors, people in those states had about four times the odds of reporting medical cannabis use as people in states without such laws. Because the research was based on observational, cross-sectional survey data, however, it cannot establish that legalization directly caused the increase. Changes in availability, public attitudes, state policy, and patterns of cannabis use may all have contributed.

The researchers also identified differences across demographic groups. Medical cannabis use was more common among people aged 18 to 25, men, people who had completed some college, those who had never married, and uninsured respondents. These findings should not be interpreted as proof that each group experienced a larger increase in use. In several cases, the statistically significant results reflected interactions between a group characteristic and the survey year rather than a simple ranking of which groups increased most.

Medical cannabis use was also associated with poorer self-rated health, a major depressive episode during the previous year, cocaine use disorder, and nonmedical use of prescription pain relievers. The authors caution that the survey cannot determine whether cannabis was recommended for these conditions, used by individuals attempting to manage them, or used alongside other treatments. It also cannot establish which condition came first.

The study did not evaluate whether cannabis helped or harmed individual patients. Its authors noted that evidence for medical cannabis varies considerably by condition. A National Academies review of cannabis and cannabinoids found evidence of benefit for some uses, including chronic pain in adults, certain multiple-sclerosis symptoms, and chemotherapy-related nausea and vomiting, while also identifying substantial gaps in knowledge and limitations in the available research.

Potential risks remain relevant even when cannabis is used for medical reasons. Cannabis can impair coordination, judgment, and reaction time, and the Centers for Disease Control and Prevention warns that cannabis-impaired driving is dangerous. Cannabis use has also been associated with psychosis and other mental-health risks, particularly among people who use it frequently or begin at a young age, although associations do not by themselves prove that cannabis caused a particular disorder.

The central finding is therefore about changing use patterns, not proven medical effectiveness. Medical cannabis use became substantially more common in the United States from 2013 through 2020, especially in states that had legalized it. Determining which products, doses, and conditions offer meaningful benefits—and understanding the long-term risks—will require better clinical studies and longer-term monitoring. Readers seeking current information about survey methods and substance-use trends can consult the official National Survey on Drug Use and Health program.