Abstract Background Bariatric surgery has been associated with altered alcohol metabolism. We examined whether patients undergoing bariatric surgery have a higher risk of developing alcohol use disorder (AUD) compared with individuals with obesity who have not received bariatric surgery. Methods In this prospective cohort study, we followed 13 430 patients undergoing bariatric surgery (95% gastric bypass) between 2005 and 2013 and a reference group of 21 021 individuals with obesity for a median of 6.9 years (5th-95th percentile: 4.0–9.8). Four different approaches were used to account for baseline differences between the two groups: (i) adjustment; (ii) inverse probability of treatment weighting (IPTW); (iii) 1:1 matching based on propensity scores; and (iv) before-and-after analysis comparing the bariatric surgery group with itself 5 years before and after surgery. Cox proportional hazard modelling was used to estimate hazard ratios of AUD defined from national registers. Results When applying the IPTW approach, the hazard ratio (HR) of AUD for bariatric surgery patients was 7.29 [95% confidence interval (CI): 5.06–9.48] compared with individuals without surgery. When employing different approaches (adjustment for baseline variables, matching on propensity scores, before-and-after analyses), results were of similar magnitude. Analysis stratified by time after surgery revealed a higher risk of AUD already within the first year following surgery [HR: 2.77 (95% CI: 1.39–5.53)]. Conclusions Patients undergoing bariatric surgery have a higher risk of developing AUD compared with individuals without bariatric surgery. The higher risk observed in this group of patients cannot be explained by differences in baseline characteristics such as socioeconomic factors. Despite the higher risk of AUD, only few individuals developed AUD. Individuals with disabling obesity should therefore not rule out surgery based on these results but rather be aware of negative implications.