Concern about the use of gonadotrophin-releasing hormone (GnRH) agonists in ovarian stimulation of poor responder IVF patients has arisen from the claim that GnRH agonists might have a direct deleterious effect through their receptors on the ovary. In this study, we compared two ovarian stimulation protocols in which no GnRH agonists were used. In all, 40 patients with a poor response in previous treatment cycles were included. They were divided into two groups: group I (n = 20) received ovarian stimulation for 20 cycles, without the addition of either GnRH agonist or antagonist; while group II (n = 20) patients received ovarian stimulation for 20 cycles, including the administration of a GnRH antagonist (Cetrorelix, 0.25 mg daily) during the late follicular phase. There was no statistically significant difference between the groups for mean age, duration of infertility, baseline FSH concentration, cancellation rate, number of ampoules of gonadotrophin used, number of mature oocytes retrieved, oestradiol concentrations on the day of injection of human chorionic gonadotrophin (HCG), fertilization rate and number of embryos transferred. The clinical pregnancy and implantation rates in group II appeared higher than in group I, but were not significantly different (20 and 13.33% compared with 6.25 and 3.44% respectively). The addition of GnRH antagonists to ovarian stimulation protocols might be a new hope for poor responder IVF patients, but this report is preliminary and further controlled randomized prospective studies with larger sample sizes are required.