OBJECTIVE To determine if premature luteinization can occur in GnRH agonist (GnRH-a) and FSH (recombinant FSH and human urinary FSH) IVF cycles and whether premature luteinization affects IVF and clinical outcome. METHODS Retrospective evaluation of 171 IVF-ET cycles. The cycles were divided into two groups according to the P level on the day of hCG: group I (serum P </= 0.9 ng/mL [conversion factor to SI unit, 3.180]) and group II (serum P >/= 1.1 ng/mL). METHODS Comparison of cycles characteristics and of cumulative exposure of follicular serum E2, FSH, LH, and P as well as of IVF and clinical outcome were made between the study groups. RESULTS Twenty-three of the 171 cycles (13.4%) demonstrated premature luteinization. The age of the patients, the E2, and LH exposure were similar between the groups. The number of the ampules of gonadotropins (recombinant FSH and urinary FSH) used and the area under FSH and P curve were higher in cycles with premature luteinization. The area under the FSH curve correlated with the area under the P curve. Similar IVF and clinical outcomes were observed in cycles with and without premature luteinization. CONCLUSIONS The greater FSH exposure and its correlation with the P exposure suggest that one of the possible factors inducing premature luteinization is the increased FSH-induced LH receptivity in granulosa cells. No adverse effects of premature luteinization on the IVF and clinical outcome were observed.