Objective: To explore factors causing a premature rise in luteinizing hormone among high ovarian responders undergoing the gonadotropin-releasing hormone (GnRH) antagonist ovarian stimulation protocol. Methods: The present retrospective study included healthy women undergoing fresh cycles using a fixed GnRH antagonist protocol with a predicted high response and antral follicle count (AFC) of at least 15 at the Reproductive Medicine Center of Tongji Hospital, China, between January 1 and December 31, 2016. Treatment-related characteristics, hormone changes, and pregnancy outcomes were compared between patients who did or did not experience a premature luteinizing hormone rise. Results: There were 314 patients included; 49 experienced premature luteinizing hormone increases. Among patients who experienced a premature rise in luteinizing hormone, a lower two pronuclear embryo rate (P=0.038); fewer high-quality embryos (P=0.020); higher serum luteinizing hormone (P=0.006), progesterone (P=0.013), and estradiol (E2) levels (P=0.003) on the day of human chorionic gonadotropin administration; a lower clinical pregnancy rate (P=0.031); and a higher cancellation rate (P=0.006) were observed. AFC of at least 22 (P=0.001) and E2 of 669 pg/mL or higher at the start of GnRH antagonist administration were predictive of early (P=0.036) and late (P=0.033) premature luteinizing hormone increases. Conclusion: Earlier administration of GnRH antagonist could avoid premature luteinizing hormone increases among high ovarian responders, especially those with a starting AFC of 22 or more.