A randomized clinical trial was performed comparing recombinant follicle stimulating hormone (rFSH, Puregon, n = 54) and human menopausal gonadotrophin (HMG, Humegon, n = 35) in infertile women undergoing in-vitro fertilization without the use of a gonadotrophin-releasing hormone (GnRH) agonist. Most patients had a tubal or idiopathic infertility, the latter always longer than 4 years' duration. Patients with sperm abnormalities were excluded. None of the between-group differences in treatment outcome was statistically significant. In the rFSH group, a mean number of 11.2 oocytes was retrieved compared with 8.3 in the HMG group. Ongoing pregnancy rates per started cycle were higher in the rFSH group (22.2%) than in the HMG group (17.1%). Implantation rates were 27.5% in the rFSH group in comparison with 16.7% in the HMG group. In the rFSH group, a mean total dose of 1410 IU during 6.2 days was administered compared with 1365 IU in 6.0 days in the HMG group. Oestradiol concentrations on the day of human chorionic gonadotrophin administration were 3889 pmol/l in the rFSH group and 3145 pmol/l in the HMG group. In 15 subjects (rFSH: n = 9, 16.7%; HMG: n = 6, 17.1%) luteinizing hormone concentrations higher than 10 IU/l were seen during stimulation. In two of them, both from the rFSH group, ongoing pregnancies were achieved. The results indicate that rFSH (Puregon) is at least as efficacious as HMG and that acceptable pregnancy rates can be achieved without the use of a GnRH agonist.