Since the introduction of gonadotropin-releasing hormone (GnRH) antagonists, an extensive amount of literature investigating the role of the downregulation protocols on pregnancy outcomes has been published. However, these studies were mainly performed in the general infertile population where patients with endometriosis were often excluded or underrepresented. This study is a large retrospective cohort study including 386 endometriosis patients undergoing IVF/ICSI, who had been previously classified according to the rAFS system. Patients were stimulated either a long GnRH agonist or GnRH antagonist protocol. Depending on endometriosis stage, patients were divided into two groups: endometriosis stage I-II and endometriosis stage III-IV. Each group was subdivided, based on the type GnRH analog used. When comparing the GnRH agonist and antagonist groups, patients with endometriosis stage I-II, had a tendency toward higher β-hCG positive, clinical pregnancy, and live birth rates (42.8% vs. 26.7%; p = .07) in favor of GnRH agonist use. In endometriosis stage III-IV, no differences were observed between agonist and antagonist cycle in any of the pregnancy outcomes. Multivariate regression analysis did not reveal any significant predictor of live birth after adjusting for relevant confounders. Based on our findings, the chance to have a liveborn in endometriosis population seems not to be affected by the type of GnRH analog used, at least in advanced stages. Findings from stage I-II endometriosis cases merit consideration and further evaluation in a larger sample size is warranted.