Purpose : The objective of this study was to compare the efficacy of GnRH-antagonists to GnRH-agonists in ovarian stimulation of poor responders undergoing IVF. Methods : Retrospective analysis of our data revealed that 56 patients underwent treatment with a GnRH-agonist according to the flare-up protocol. Patients failing to achieve an ongoing pregnancy (n = 53) were subsequently treated in the next cycle with a GnRH-antagonist according to the multiple-dose protocol. Main outcome measures included the clinical pregnancy and implantation rates. Results : While ovulation induction characteristics and results did not differ between the two protocols, the number of embryos transferred was significantly higher (P = 0.046) in the GnRH-antagonist than in the GnRH-agonist stimulation protocol (2.5 ± 1.6 vs. 2.0 ± 1.4, respectively). The clinical pregnancy and implantation rates per transfer in the GnRH-antagonist group appeared higher than in the GnRH-agonist, but did not differ statistically (26.1 and 10.7 compared with 12.2 and 5.9%, respectively). However, the ongoing pregnancy rate per transfer was statistically significantly higher (P = 0.03) in the GnRH-antagonist than in the GnRH-agonist group (23.9 vs. 7.3%, respectively). Conclusion : Applying GnRH-antagonists to ovarian stimulation protocols may offer new hope for IVF poor responder patients. However, further controlled randomized prospective studies with larger sample sizes are required to establish these results.