OBJECTIVE: The purpose of this study was to identify useful clinical factors for the identification of patients with polycystic ovary syndrome (PCOS) who would benefit from in vitro maturation (IVM) treatment without exhibiting compromised pregnancy outcomes. METHODS: A retrospective cohort study was performed of 186 consecutive patients with PCOS who underwent human chorionic gonadotropin-primed IVM treatment between March 2010 and March 2014. Only the first IVM cycle of each patient was included in this study. A retrospective case-control study was subsequently conducted to compare pregnancy outcomes between IVM and conventional in vitro fertilization (IVF) cycles. RESULTS: Through logistic regression analyses, we arrived at the novel finding that serum anti-Müllerian hormone (AMH) levels and the number of fertilized oocytes in IVM were independent predictive factors for live birth with unstandardized coefficients of 0.078 (95% confidence interval [CI], 1.005-1.164; p=0.037) and 0.113 (95% CI, 1.038-1.208; p=0.003), respectively. Furthermore, these two parameters were able to discriminate patients who experienced live births from non-pregnant IVM patients using cut-off levels of 8.5 ng/mL and five fertilized oocytes, respectively. A subsequent retrospective case-control study of patients with PCOS who had serum AMH levels ≥8.5 ng/mL showed that IVM had pregnancy outcomes comparable to conventional IVF, and that no cases of ovarian hyperstimulation syndrome were observed. CONCLUSION: Serum AMH levels are a useful factor for predicting pregnancy outcomes in PCOS patients before the beginning of an IVM cycle. IVM may be an alternative to conventional IVF for PCOS patients if the patients are properly selected according to predictive factors such as serum AMH levels.