OBJECTIVE: To evaluate the association between cannabis use and testicular function (as assessed through semen quality and serum hormone levels) in different populations. EVIDENCE REVIEW: Systematic review and meta-analysis of population-based retrospective cohort studies. PRISMA guidelines were used for abstracting data and assessing data quality and validity. Data were pooled using a fixed-effects or random-effects model depending on the heterogeneity of studies included. Pooled risk ratio (RR) of having any sperm abnormality and testosterone, FSH, and LH standardized mean differences among male cannabis users and non-users, and meta-regression analysis according to age and year of publication. RESULTS: Nine studies were evaluated which included 4014 men with semen data and 4787 with hormonal data. Overall among 1158 cannabis users, 44.9% had impaired semen parameters, compared with 24.5% of the 2856 non-users. The relative risk among cannabis users for any abnormal semen parameter was 1.159 (95% CI: 0.840; 1.599, P = 0.369). The standardized mean difference between user and non-user testosterone levels was -0.139 (95% CI: -0.413; 0.134, P = 0.318). For FSH, the standardized mean difference estimate was -0.142 (95% CI: -0.243; -0.0425, P = 0.005), while for LH the standardized mean difference estimate was -0.318 (95% CI: -0.810-0.175; P = 0.206). CONCLUSIONS: The current evidence does not suggest clinically significant associations between cannabis use and testicular function. However, we cannot exclude an effect of cannabis because of the limited and heterogeneous studies. Additionally, well-designed studies will be needed to define the association between cannabis use and the male reproductive system.