We studied 15 anovulatory women undergoing ovulation induction with purified human urinary FSH or purified human urinary FSH and LH [human menopausal gonadotropins (hMG)]. All patients had either sporadic or no vaginal bleeding after progesterone therapy and failed to ovulate after receiving clomiphene (250 mg for 5 days) plus hCG. Other causes of infertility were ruled out. Sixteen cycles of FSH and 12 cycles of hMG were administered according to a standard protocol. Estradiol, progesterone, androstenedione, testosterone, LH, and FSH concentrations were quantitated by RIA. Follicular diameter was determined using ultrasound. There was no significant difference in the amount of FSH or hMG used per patient, in the duration of therapy before hCG administration, or in the length of the luteal phase in any patient. There was a difference in the number of follicles greater than 1000 mm3 per cycle in those patients receiving FSH compared to the number in those receiving hMG [2.8 +/- 1.3 (+/- SEM) vs. 4.4 +/- 1.5 follicles; P = 0.026). The maximum follicular phase serum estradiol (18.3 vs. 34.8 ng/ml) and maximum luteal phase progesterone concentrations (1289 vs. 2808 pg/ml; P = 0.026) were also different between the FSH and hMG groups. Linear regression analysis revealed a significant correlation between the peripheral serum estradiol levels and the total follicular volume of follicles in the hMG-treated group which was not apparent in the FSH-treated group. These findings suggest that exogenous LH may not be required to induce folliculogenesis in anovulatory patients.