Abstract Study question Comparison of pregnancy and perinatal outcomes in blastocysts with poor inner cell mass quality or poor trophectoderm quality to optimize non-quality single blastocyst transfer strategies. Summary answer Poor trophoblast quality blastocysts have better pregnancy outcomes than poor inner cell mass quality blastocysts, with no statistically significant difference in perinatal outcomes What is known already There is no doubt that good-quality blastocysts have better pregnancy outcomes than poor-quality blastocysts, but the preference for trophectodermal poor-quality blastocysts or inner cell mass poor-quality blastocysts in non-good-quality blastocyst cycles is still uncertain. Study design, size, duration A total of 46,433 first single blastocyst transfer cycles from March 1, 2014, to December 31, 2022 were retrospectively analyzed, and blastocyst transfer cycles were divided into 3 groups according to the grade of trophoderm and inner cell mass: A/B A/B (Group 0, N = 30748), A/B C (Group 1, N = 15455), and C A/B (Group 2, N = 230). Participants/materials, setting, methods 46,433 patients who underwent their first single blastocyst transfer at Reproductive and Genetic Hospital of CITIC-Xiangya were included. Multivariate log-binomial regression model with COPY methods were performed to estimate relative risk of Group 2/1 versus Group 0 and Group 2 versus Group 1 for reproductive and neonatal outcomes. Five propensity score weighting methods, namely IPW, processing, overlap, matching, and entropy, were used for sensitivity analysis to explore the robustness of the results. Main results and the role of chance After adjusting for female age, BMI, duration of infertility, number of cycles, AFC, AMH, endometrial thickness, causes of infertility (tubal disorders, ovulatory disorders, endometriosis, unexplained, more than one causes, male factors), culture days (D5, D6), expansion of blastocyst (3, 4, 5, 6), and type of blastocyst (fresh blastocyst, thawed blastocyst, thawed cleavage cultured to blastocyst, PGT blastocyst), we found that clinical pregnancy rate (CPR) and live birth rate (LBR) of Group 0, Group 1 and Group 2 were successively significantly decreased (CPR: Group 1 vs Group 0, RR = 0.83 (0.81, 0.84), Group 2 vs Group 0, RR = 0.43 (0.35, 0.53), Group 2 vs Group 1, RR = 0.53 (0.42, 0.64); LBR: Group 1 vs Group 0, RR = 0.79 (0.77, 0.81), Group 2 vs Group 0, RR = 0.36 (0.27, 0.47), Group 2 vs Group 1, RR = 0.45 (0.34, 0.59); respectively), while early miscarriage rate was successively significantly increased (Group 1 vs Group 0, RR = 1.3 (1.22, 1.38), Group 2 vs Group 0, RR = 2.25 (1.46, 3.15), Group 2 vs Group 1, RR = 1.73 (1.12, 2.42)). Late miscarriage, preterm birth, singleton live birth, sex ratio of singleton, low birth-weight of singleton, and birth weight of singleton were not statistically different between three groups. Above results were robust in our two sensitivity analysis. Limitations, reasons for caution As a retrospective study, there are inevitably unmeasured confounding factors. This study only included a small number of Group 2. It’s therefore possible that our study was under powered. Risk of gestational hypertension, gestational diabetes, cesarean section and macrosomia should also be evaluated with a more comprehensive assessment of safety. Wider implications of the findings High-quality blastocysts have better pregnancy and live birth rates than poor-quality blastocysts, but there are no significant differences in preterm birth, sex ratio, or birth weight. In the absence of good quality blastocysts for transplantation, poor quality blastocysts of the trophectoderm are preferred over poor quality inner cell mass. Trial registration number not applicable