Problems arising from controlled ovarian hyperstimulation for intrauterine insemination, such as premature luteinization and asynchronous ovarian follicular development, are identical to those encountered with controlled ovarian hyperstimulation for in vitro fertilization (IVF) and gamete intrafallopian transfer (GIFT). It has been suggested that the adjunctive use of GnRH agonists for controlled ovarian hyperstimulation improves the efficiency of GIFT and IVF cycles. We hypothesized that adjunctive use of leuprolide acetate, a GnRH agonist, would have a similarly beneficial effect on cycle quality and cycle fecundity in subfertile women treated with controlled ovarian hyperstimulation and intrauterine insemination. We randomly assigned the first cycle of controlled ovarian hyperstimulation and intrauterine insemination for each of 97 subfertile women to include either human menopausal gonadotropins (hMGs) alone or hMGs following midluteal pre-treatment with leuprolide. If a pregnancy did not occur in the first cycle, the woman was given the other treatment in the second cycle. Although the cycles that included leuprolide required a larger amount of hMGs and more days of stimulation per cycle, the mean estradiol concentrations and numbers of follicles were not different. Despite prevention of premature luteinization with leuprolide, the cycle fecundity was not different between groups (0.11 with adjunctive leuprolide treatment and 0.22 with hMGs alone). We conclude that in unselected subfertile patients, the adjunctive use of leuprolide for controlled ovarian hyperstimulation and intrauterine insemination does not improve cycle fecundity compared with treatment cycles that do not include adjunctive leuprolide therapy.