OBJECTIVE To compare the use of GnRH agonist (GnRH-a) versus hCG in triggering the follicular rupture in patients with polycystic ovarian disease (PCOD) in whom ovulation was induced by gonadotropins. METHODS Polycystic ovarian disease gonadotropin hyperstimulated cycles outcome was investigated in a prospective study. METHODS Thirty-three PCOD patients (40 cycles) with gonadotropin-induced mild to moderate degree of ovarian hyperstimulation received 5,000 IU IM hCG or 200 microg [corrected] SC GnRH-a. A subgroup of GnRH-a-treated patients received P for luteal support. Five GnRH-a-treated patients underwent a GnRH test during luteal phase. METHODS Echographic and endocrine characteristics both during the therapy and the luteal phase. RESULTS There was a similar percentage of ovulation and pregnancy rate in both groups of patients. The ovarian enlargement during the luteal phase in the GnRH-a-treated patients was lower than in the hCG group. Progesterone plasma levels (at midluteal phase) and the length of luteal phase was significantly lower in GnRH-a-treated patients with respect to the hCG-treated group. These differences disappeared in patients receiving luteal support. After GnRH injection, LH secretion decreased in GnRH-a-treated patients with respect to controls; however, corpus luteum was able to respond with a normal increase of P production. CONCLUSIONS The GnRH-a appears to be an effective alternative to hCG for inducing the follicular rupture in stimulated cycles in women who are at risk for developing ovarian hyperstimulation syndrome. However, GnRH-a administration can induce short luteal phase. This defect may be ascribed to the pituitary desensitization rather than to a direct effect on corpus luteum. Luteal phase support is needed to prevent luteal phase deficiency.