The reports on how to stimulate the ovaries for oocyte retrieval in good prognosis patients are contradictory and often favor one type of controlled ovarian hyperstimulation (COH). For this reason, we retrospectively analyzed data from IVF/ICSI cycles carried out at our IVF Unit in good prognosis patients (aged <38 years, first and second attempts of IVF/ICSI, more than 3 oocytes retrieved) to elucidate which type of COH is optimal at our condition. The included patients were undergoing COH using GnRH agonist, GnRH antagonist or GnRH antagonist mild protocol in combination with gonadotrophins. We found significant differences in the average number of retrieved oocytes, immature oocytes, fertilized oocytes, embryos, transferred embryos, embryos frozen per cycle, and cycles with embryo freezing between studied COH protocols. Although there were no differences in live birth rate (LBR), miscarriages, and ectopic pregnancies between compared protocols, pregnancy rate was significantly higher in GnRH antagonist mild protocol in comparison with both GnRH antagonist and GnRH agonist protocols and cumulative LBR per cycle was significantly higher in GnRH antagonist mild protocol in comparison to GnRH agonist protocol. Our data show that GnRH antagonist mild protocol of COH could be the best method of choice in good prognosis patients.