Introduction Prior research has examined the prevalence and correlates of cannabis use disorder (CUD) in people who use cannabis; however, these are poorly described for people using cannabis for medical reasons. Methods Data came from a 2018 to 2019 online, anonymous, cross‐sectional survey of Australians reporting using either illicit or licit cannabis for medical reasons within the past year. Included were questions on demographics, current and lifetime patterns of cannabis use, clinical conditions for which medical cannabis was used, and individual criteria for CUD and cannabis withdrawal syndrome. Bayesian Horseshoe logistic regression models were used to identify covariates associated with meeting CUD DSM‐5 conditions for any‐CUD (≥2/11 criteria) and moderate–severe‐CUD (≥4/11). Results A total of 905 participants were included in the analysis. The majority (98%) used illicit cannabis products. Criteria for any‐CUD criteria were met by 290 (32.0%), and 117 (12.9%) met criteria for moderate–severe‐CUD. Tolerance (21%) and withdrawal (35%) were the most commonly met criteria. Correlates with the strongest association with CUD were inhaled route of administration [odds ratio (OR) = 2.96, 95% credible interval 1.11, 7.06], frequency of cannabis use (OR = 1.24, 1.11–1.35), proportion of cannabis for medical reasons (OR = 0.83, 0.74, 0.94), frequency of tobacco use (OR = 1.10, 1.03, 1.17), age (OR = 0.75, 0.64, 0.90) and pain as main clinical indication (OR = 0.58, 0.36, 1.00). Discussion and Conclusions Prevalence of CUD in medical cannabis users appears comparable to ‘recreational’ users, with many similar correlates. CUD was associated with using cannabis to treat mental health rather than pain conditions and inhaled over other routes of administration.

No abstract