Two hundred ten treatment cycles of follicle-stimulating hormone (FSH) or human menopausal gonadotropin (hMG) were completed in 49 patients with clomiphene citrate-resistant polycystic ovarian syndrome. The results from 68 cycles of daily intramuscular (IM) FSH and 41 cycles of IM hMG were compared. The ovulation rate, maximum serum estradiol (E2) levels achieved, and pregnancy rate were similar in both groups, but FSH resulted in significantly fewer follicles developing and hyperstimulation. The 68 cycles of daily IM FSH were further compared with the outcome of administering the FSH as an alternate-day IM injection in 70 cycles, and by subcutaneous pulsatile injection in 31 cycles. There were no differences in any of the parameters measured between daily and alternate-day FSH. Pulsatile FSH required a greater total dose over a longer period of time to achieve stimulation. It also produced fewer follicles, a lower maximum serum E2 level, and the lowest incidence of hyperstimulation. Twenty pregnancies resulted, of which 6 aborted in the first trimester; there was 1 set of twins and 13 singleton pregnancies. The cumulative pregnancy rate after 6 treatment cycles was 62%.