Polycystic ovarian syndrome (PCOS) is a complex disorder with multiple abnormalities, including hyperandrogenism, ovulatory dysfunction, and altered gonadotropin secretion. The majority of patients have elevated LH levels in plasma and a persistent rapid frequency of LH (GnRH) pulse secretion, the mechanisms of which are unclear. Earlier work has suggested that the sensitivity of the GnRH pulse generator to inhibition by ovarian steroids is impaired. We performed a study to determine whether antiandrogen therapy with flutamide could enhance feedback inhibition by estradiol (E2) and progesterone (P) in women with PCOS. Ten anovulatory women with PCOS and nine normal controls (days 8-10 of the cycle) were studied on three occasions. During each admission, LH and FSH were determined every 10 min and E2, P, and testosterone (T) every 2 h for 13 h. After 12 h, GnRH (25 ng/kg) was given iv. After the first admission, patients were started on flutamide (250 mg twice daily), which was continued for the entire study. The second admission occurred on days 8-10 of the next menstrual cycle for normal controls and on study day 28 for PCOS patients. Subjects were then given E2 transdermally (mean plasma E2, 106+/-18 pg/mL) and P by vaginal suppository to obtain varied plasma concentrations of P (mean P, 4.4+/-0.5 ng/mL; range, 0.6-9.0 ng/mL), and a third study was performed 7 days later. At baseline women with PCOS had higher LH pulse amplitude, response to GnRH, T, androstenedione, and insulin and lower sex hormone-binding globulin concentrations (P < 0.05). Most hormonal parameters were not altered by 4 weeks of flutamide, except T in controls and E2 and FSH in PCOS patients, which were lower. Of note, flutamide alone had no effect on LH pulse frequency or amplitude, mean plasma LH, or LH responsiveness to exogenous GnRH. After the addition of E2 and P for 7 days, both PCOS patients and normal controls had similar reductions in LH pulse frequency (4.0+/-0.7 and 5.8+/-0.7 pulses/12 h, respectively). This contrasts with our earlier results in the absence of flutamide, where a plasma P level of less than 10 ng/mL had minimal effects on LH pulse frequency in women with PCOS, but was effective in controls. These results suggest that although the elevated LH pulse frequency in PCOS may in part reflect impaired sensitivity to E2 and P, continuing actions of hyperandrogenemia are important for sustaining the abnormal hypothalamic sensitivity to feedback inhibition by ovarian steroids.