Introduction Progestins are commonly prescribed first‐line drugs for endometriosis. High rates of non‐response and intolerance to these drugs have been previously reported. However, no study to date has investigated the characteristics and comorbidities of patients taking progestins in relation to treatment outcomes, so identifying which patients will respond to or tolerate the treatment is currently impossible. The purpose of this study, therefore, was to identify risk factors for non‐response and discontinuation of Dienogest (DNG) in women with endometriosis. Material and methods This is a retrospective cohort study including women currently taking, or newly prescribed, DNG for endometriosis‐associated pain presenting in the Endometriosis Clinic of the University Hospital of Bern between January 2017 and May 2018. Women with initiation of treatment directly after surgery for endometriosis were excluded. For all participants the symptoms and comorbidities were documented. Effectiveness, tolerability and discontinuation of DNG were the primary end points. Univariate and multivariate binary logistic regression models were carried out to identify risk factors for non‐response, intolerance and discontinuation of DNG. Results A sufficient or excellent treatment response was reported by 85/125 (68%) participants. Genital bleeding during the DNG treatment was negatively (OR 0.185, 95% CI 0.056–0.610, P = .006) and rASRM endometriosis stages III and IV were positively (OR 3.876, 95% CI 1.202–12.498, P = .023) correlated with the DNG response. When accounting for exclusively pretreatment factors, primary dysmenorrhea (OR 0.236, 95% CI 0.090–0.615, P = .003) and suspicion of adenomyosis (OR 0.347, 95% CI 0.135–0.894, P = .028) were inversely correlated with DNG response, and the latter was also correlated with treatment discontinuation (OR 3.189, 95% CI 1.247–8.153, P = .015). Conclusions Genital bleeding during the DNG treatment and low rASRM stages are independent risk factors for DNG non‐response. Before treatment initiation, primary dysmenorrhea and suspicion of adenomyosis correlate with DNG non‐response. The results could assist the clinician first to provide detailed information to women before treatment initiation, second to identify and possibly modify in‐therapy factors correlated to treatment effectiveness and lastly to switch treatment on time if needed.