A potent GnRH antagonist, [Ac-delta 3-Pro1,p-F-D-Phe2,D-Trp3,6]GnRH (4F-Antag), was used as a probe to determine the relative dependency of ovarian cyclicity on pulsatile gonadotropin secretion. 4F-Antag was administered in a dose of 80 micrograms/kg sc twice a day for 3 consecutive days during different phases of the menstrual cycle. This treatment resulted in a prompt attenuation of pulsatile gonadotropin secretion in all women studied. Maximal suppression of gonadotropin levels, expressed as percent change from baseline, averaged 48% for LH and 22% for FSH. The reduced pulsatile gonadotropin release induced by 4F-Antag administration during the early follicular phase resulted in a significant decrease in serum estradiol levels during the period of treatment and was followed by a prolongation of follicular phase (2.4 days) and cycle length (3.5 days), but no alteration of subsequent cyclic ovarian steroid profiles compared to control cycles. Treatment initiated during the midfollicular phase 4-6 days before the expected LH surge resulted in a more dramatic decline in serum estradiol levels and prolongation of follicular phase length by 5-6 days compared to control cycles. Normal luteal function was preserved. These alterations were compatible with induction of the demise of the dominant follicle followed by the reinitiation of follicular recruitment. Administration of 4F-Antag during the midluteal phase resulted in rapid falls in serum estradiol and progesterone levels and the onset of menstrual bleeding in all women. The luteolytic effect of 4F-Antag was completely negated by the administration of hCG. These data indicate that 4F-Antag interferes with ongoing cyclic ovarian function by reducing pulsatile gonadotropin stimulation, which disrupts folliculogenic processes and induces the demise of the corpus luteum. From these findings we infer that the functional integrity of ovarian cyclicity is remarkably sensitive to brief (3 days) and partial reduction in pulsatile gonadotropin secretion.