Abstract Study question What are the reproductive outcomes and the prognostic factors of live birth rates in endometriosis patients referred to oocyte donation after multiple in vitro fertilization(IVF)-failures? Summary answer In this specific population, oocyte donation appears to be a viable option to optimize the reproductive outcomes, as the cumulative live birth rate reaches 63.2%. What is known already Assisted Reproductive Technology(ART) is currently recognized as one of the main therapeutic options to manage endometriosis-related infertility, often resulting in satisfactory outcome. A burning issue in this field remains how to deal with patients undergoing recurrent ART failures. As there appear to be two main factors contributing to the failures, i.e., a deleterious effect of endometriosis on embryo implantation and altered oocyte and embryo quality, two competing therapeutic strategies can be considered:(i) endometriosis surgery and (ii) oocyte donation. However, no data has been published to date in the specific population of endometriosis patients with previous IVF failures, undergoing oocyte donation. Study design, size, duration Observational cohort study including fifty-seven women with endometriosis-related infertility and multiple IVF failures (≥ 2 failed IVF cycles) referred to oocyte donation who consulted at our institution between January 2013 and June 2022. Participants/materials, setting, methods Endometriosis was diagnosed based on published imaging criteria using transvaginal sonography and magnetic resonance imaging and confirmed histologically in women who had a history of surgery for endometriosis. The main outcomes measured were the clinical pregnancy rate (CPR) and the cumulative live birth rate (CLBR). We compared the characteristics of women who had a live birth versus those who did not using univariate and multivariate analysis to identify determinant factors of fertility outcome. Main results and the role of chance Fifty-seven patients underwent 90 oocyte donation cycles, after 244 failed autologous IVF cycles. The mean age of the population was 36.8 ± 3.3 years, with a mean infertility duration of 3.6 ± 2.2 years, and a mean number of autologous IVF/ICSI cycles of 4.4 ± 2.3 cycles per patient. Three patients (5.3%) had superficial endometriosis, 2 patients (3.5%) had ovarian endometriomas, and 52 patients (91.2%) had deep infiltrating endometriosis, among which 30 patients (57.7%) had bowel lesions. Thirty patients (52.6%) had associated adenomyosis. Overall, the CPR and the CLBR per patient were 49/57 (85.9%) and 36/57 (63.2%), respectively. After multivariate analysis, only being nulligravida (p = 0.002) remained an independent negative predictive factor of the live birth rate. Limitations, reasons for caution In 35.1% of patients there was no surgical/histological proof of endometriosis. However, for all the study participants, endometriosis was diagnosed and staged using patient interviews, clinical examination, and imaging techniques to assess for endometriosis phenotypes, which is in line with the structured diagnostic process acknowledged as being appropriate for endometriosis. Wider implications of the findings In this specific population, poor embryo quality may be the main limiting factor of ART success rates, rather than impaired endometrial receptivity. These data are reason for clinicians not to systematically refer infertile endometriosis patients with recurrent IVF failures to surgery. Trial registration number not applicable