In primates, the LH surge that triggers ovulation is induced by an increase in circulating estradiol (E(2)) levels. Although several studies suggest that E(2) acts on the pituitary, it is still not clear whether GnRH is involved. We investigated the role of GnRH during the periovulatory period in normal women by treating them with the GnRH antagonist Nal-Glu ([Ac-D2Nal(1),D4-CIPhe(2),D3pal(3),Arg(5),DGlu(6) (AA),DAla(10)] when E(2) levels exceeded 550 pmol/L. In the first study (A), Nal-Glu was administered in five regimens (n = 4 in each group): a single sc injection of 10 mg (group 1), a single injection of 20 mg (group 2), and an injection of 10 mg, sc, on 2 (group 3), 3 (group 4), and 5 consecutive days (group 5). In the second study (B; n = 4), Nal-Glu (10 mg, sc, on 3 consecutive days) was coadministered with Eg benzoate (EB; 0.5 mg, im, every 12 h on 3 consecutive days). Controls (n = 4) were treated with EB alone at the same stage of the cycle. In the third study (C), three women received 10 mg/day Nal-Glu, sc, on 3 consecutive days together with pulsatile GnRH therapy (25 mu g/pulse, one pulse every 90 min, sc, for 3 days); the first pulse was given 12 h after the first NalGlu injection. In study A, gonadotropin suppression resulted in a transient decline in Ep in groups 1 and 2. Relative to control cycles, the LH surge occurred with a delay of 24-48 h in group 1 and 24-120 h in group 2. In groups 3, 4, and 5, Nal-Glu administration resulted in the demise of the dominant follicle in half of the women in each group. The remaining women showed a profile similar to that of groups 1 and 2, Le. a transient decline in E(2) levels followed by a recovery, and a LH surge occurring 4 +/- 0.3 days after the last Nal.Glu injection. In study B, simultaneous administration of Nal-Glu and EB induced a rise in EP levels from 951.3 +/- 79.6 to 4000.1 + 772.5 pmol/L 24 h after the beginning of treatment. Serum LH and FSH levels both decreased and remained low throughout Nal-Glu treatment. None of the women showed a LH rise in response to the EB injection. In controls, however, EB administration was followed by a significant gonadotropin discharge 48 h after the first EB injection. In study C, coadministration of Nal-Glu with pulsatile GnRH therapy resulted in a rapid increase in EO concentrations, leading to apparently normal preovulatory Ee levels preceding the LH surge occurring during Nal-Glu treatment. In conclusion, administration of a GnRH antagonist during the periovulatory period prevents the LH surge and suppresses the positive feedback of EB. Our findings show that in women, GnRH secretion is required to start the estrogen-induced LH surge.