Background: Cannabis use by physicians can be detrimental for them and their patients. We conducted a systematic review and meta-analysis on the prevalence of cannabis use by medical doctors (MDs)/students. Method: PubMed, Cochrane, Embase, PsycInfo and ScienceDirect were searched for studies reporting cannabis use in MDs/students. For each frequency of use (lifetime/past year/past month/daily), we stratified a random effect meta-analysis depending on specialties, education level, continents, and periods of time, which were further compared using meta-regressions. Results: We included 54 studies with a total of 42,936 MDs/students: 20,267 MDs, 20,063 medical students, and 1976 residents. Overall, 37% had used cannabis at least once over their lifetime, 14% over the past year, 8% over the past month and 1.1 per thousand (‰) had a daily use. Medical students had a greater cannabis use than MDs over their lifetime (38% vs. 35%, p < 0.001), the past year (24% vs. 5%, p < 0.001), and the past month (10% vs. 2%, p < 0.05), without significance for daily use (0.5% vs. 0.05%, NS). Insufficient data precluded comparisons among medical specialties. MDs/students from Asian countries seemed to have the lowest cannabis use: 16% over their lifetime, 10% in the past year, 1% in the past month, and 0.4% daily. Regarding periods of time, cannabis use seems to follow a U-shape, with a high use before 1990, followed by a decrease between 1990 and 2005, and a rebound after 2005. Younger and male MDs/students had the highest cannabis use. Conclusions: If more than a third of MDs tried cannabis at least once in their lifetime, this means its daily use is low but not uncommon (1.1‰). Medical students are the biggest cannabis users. Despite being common worldwide, cannabis use is predominant in the West, with a rebound since 2005 making salient those public health interventions during the early stage of medical studies.