It has been demonstrated previously that administration of the antiprogestin mifepristone (RU 486; 1-5 mg daily) inhibits or delays both the pre-ovulatory luteinizing hormone (LH) surge and ovulation. To investigate this mechanism, dynamic tests of pituitary ovarian function were performed in six healthy women before and during the administration of mifepristone (2 mg daily for 30 days). On day 9 of the control and treatment cycles, samples of blood were collected every 15 min over 12 h for measurement of LH concentration. After 10 h, the responsiveness of the pituitary was tested by the i.v. injection of 10 micrograms of gonadotrophin-releasing hormone (GnRH). On day 10 of the control and treatment cycles, two patches releasing 200 micrograms/day of oestradiol were applied to skin on the abdomen for 3 days. Blood was collected at 24, 48, 59, 72, 81 and 96 h after application of the oestrogen patches for the measurement of gonadotrophin and ovarian hormone concentrations. Follicular development continued in all women during their treatment with mifepristone, and ovulation was suppressed (four women) or delayed (two women). There was no significant difference in the basal concentration of LH between the control and treatment cycles (mean +/- SE; 5.5 +/- 0.4 versus 7.7 +/- 0.4 IU/l respectively), or in the frequency (interpulse interval, 101 +/- 12 versus 105 +/- 13 min respectively) and the amplitude (2.1 +/- 0.4 versus 2.6 +/- 0.4 IU/l respectively) of LH pulses. The response to GnRH was similar. On day 10, the basal concentrations of LH, follicle-stimulating hormone (FSH), prolactin, oestradiol and progesterone and the diameter of the dominant follicle (15.7 +/- 1.8 versus 13.3 +/- 1.9 mm) were similar during control and treatment cycles. In control cycles, there were significant increases in the concentrations of LH and FSH within 72 h of application of the oestrogen patches. During treatment cycles, concentrations of FSH and LH remained low, and were significantly lower than the values observed during control cycles (P < 0.006). We conclude that the antiprogestin mifepristone disrupts ovulation by inhibiting the positive feedback effect of oestrogens and, hence, prevents or delays the generation of a pre-ovulatory LH surge.