Abstract Study question Does difference exist in the outcomes of in vitro fertilization and embryo transfer (IVF-ET) among ethnic Tibetan, Yi and Han Chinese women? Summary answer We report that the outcomes of IVF-ET were poorer among ethnic Yi women compared with ethnic Han and Tibetan women. What is known already It has been reported that the outcomes of IVF-ET are correlated with ethnicity/race in Europe and the United States as an independent factor. Sichuan is the only province in China where all 56 ethnic groups reside, with ethnic Han, Yi and Tibetan Chinese having the highest proportions. In Chengdu city (the capital of Sichuan province, basin), the majority ethnic group is Han, whilst in Liangshan and Aba Prefectures (plateau), the major ethnic groups are Yis and Tibetans. The living environment, diet, custom, medical condition and genetic backgrounds are different among different ethnic groups. Study design, size, duration In this multi-center retrospective cohort study, a total of 58374 Chines women aged 20 ∼ 45 had undergoing the first IVF or intracytoplasmic sperm injection (ICSI) ET treatment cycle with non-donor gametes from January 2014 to December 2022 were enrolled. Participants/materials, setting, methods The patients were enrolled from two University Affiliated Hospitals and one Reproductive Hospital in Chengdu. The controlled ovarian stimulation protocols were gonadotropin releasing hormone (GnRH) agonist protocol or GnRH antagonist protocol. The patients were divided into three groups based on their ethnic origins: Tibetans (n = 1069), Yis (n = 2273) and Hans (n = 55032). The outcomes of IVF-ET were compared, with the primary outcome being the live birth rate. Main results and the role of chance The proportion of GnRH agonist protocol and IVF fertilization, the rates of moderate and severe ovarian hyper-stimulation syndrome (OHSS), and cleavage embryo formation in ethnic Han women were significantly higher than in the ethnic Yis and Tibetans (P < 0.05). The rates of implantation, clinical pregnancy and live birth (35.5% vs. 42.4%) were significantly lower, whereas the late abortion rate was significantly higher in Yi women than that in Han women (P < 0.05). The live birth rate was significantly lower in Yi women (35.5%) than in Tibetan (43.3%) and Han women (42.2%) (P < 0.05). Adjustment was made for age, duration, type and etiology of infertility, history of spontaneous abortion, smoking, alcohol, tuberculosis infection, body mass index, antral follicle count, antimüllerian hormone, and so on, by multivariate logistic regression analysis. Yi women had significantly lower rates of moderate and severe OHSS (OR = 0.69, 95% CI = 0.49-0.94, P = 0.024) and clinical pregnancy (OR 0.8, 95% CI 0.70-0.93, P = 0.002) compared with ethnic Hans. Yi women had significantly lower live birth rate compared with ethnic Hans (OR = 0.79, 95% CI = 0.68-0.92, P = 0.002) and Tibetans (OR 0.73, 95% CI 0.58-0.92, P = 0.007). Limitations, reasons for caution The limitations are related with the retrospective study and lack of some demographic characteristics of the patients which may influence the outcomes of IVF-ET, including diet habits, educational level, and economic status. Wider implications of the findings The ethnic group is an independent factor for the outcomes of IVF-ET among ethnic Tibetan, Yi and Han women. Among the three ethnic groups, the Yis had lower live birth rate and moderate and severe OHSS, whilst ethnic Han women had more moderate and severe OHSS. Trial registration number ChiCTR2300070269