OBJECTIVE To compare the efficacy of highly purified human urinary follicle stimulating hormone (HP-hFSH) versus human recombinant follitropin-alpha (rFSH) in volunteers undergoing controlled ovarian stimulation for IVF. METHODS A randomized, controlled, investigator-blind trial. METHODS Four assisted reproductive technology centers. METHODS One hundred fifty-two IVF patients. METHODS Volunteers, aged 18-39, were randomized to HP-hFSH (n = 76) versus rFSH (n = 76) at a starting dose of 300 IU in down-regulated cycles. METHODS Number of oocytes, clinical pregnancy rate, and live birth rate with HP-hFSH versus rFSH. RESULTS The total IU of gonadotropin used did not differ between the two groups. There was no difference in number of oocytes retrieved with HP-hFSH (mean = 16.3) compared with rFSH (mean = 17.1), confidence interval (CI) of difference = -3.79 to +2.18. Clinical pregnancy rate, as defined by the presence of a gestational sac, was 48.7% (CI = 37.0%-60.4%) with HP-hFSH versus 44.7% (CI = 33.3%-56.6%) with rFSH (CI of difference = -11.9% to +19.8%). Live birth rate was 38.2% (29 of 76) in both groups (CI = 27.2%-50.0%), for a difference between groups of 0.0% (CI of the difference = -15.4% to +15.4%). CONCLUSIONS There were no statistically significant differences in mean oocyte number, clinical pregnancy rate, or live birth rate between HP-hFSH versus rFSH.