BackgroundMechanical endometrial injury prior to IVF has been suggested as a means to increase implantation rates by improving endometrial receptivity. However, the effects of endometrial injury in proliferative vs. luteal phase have not been studied before. This study aimed to explore whether endometrial injury in the proliferative phase of the preceding cycle before in vitro fertilization/embryo transfer (IVF-ET) improves the clinical outcomes in unselected subfertile women compared with injury in luteal phase.MethodsA group of 142 patients who were good responders to hormonal stimulation were randomized into four groups: injury group (group A: endometrial injury in proliferative phase, n = 38; group B: endometrium injury in luteal phase, n = 32), and non-injury group as control (group C: non-injury in proliferative phase, n = 36; group D: non-injury in luteal phase, n = 36). Patients in injury groups underwent endometrial injury in either proliferative phase or luteal phase in the preceding cycle before IVF treatment. Clinical outcomes including implantation, pregnancy, and live birth rates were analyzed among the four groups.ResultsThe baseline characteristics of the four groups including age, body mass index, duration, type and causes of infertility were similar. There were no significant differences in implantation, clinical pregnancy or live birth rates between injury group and non-injury group. Moreover, there were also no significant differences in implantation, clinical pregnancy, or live birth rates in injury in proliferative phase compared with luteal phase.ConclusionsEndometrial injury in the cycle preceding IVF of unselected subfertile women does not increase implantation, clinical pregnancy, or live birth rates. Furthermore, there is no significant difference in clinical outcomes between endometrial injury in the proliferative phase and injury in the luteal phase.Trial registrationThis study was retrospectively registered on May 26th, 2017 (ChiCTR-IOR-17011506).