Résumé : This study is a long-term evaluation of the total pregnancy potential of cohorts of fresh and cryopreserved sibling embryos from in-vitro fertilization (IVF) cycles stimulated with either the gonadotrophin-releasing hormone analogue buserelin (BUS) (long protocol) or clomiphene citrate (CC) both in combination with human menopausal gonadotrophin (HMG). Therefore a retrospective analysis was performed on patients who entered the IVF programme between January 1986 and July 1987 and who had triple embryo transfer in the collection cycle. Significantly more fertilized oocytes developed to good-quality embryos in the CC-HMG group (86.1%) than in the BUS-HMG group (80.8%). Transfer of the three morphologically best-looking embryos was performed on day 2 post-insemination in 106 CC-HMG and 80 BUS-HMG cycles. Supernumerary embryos were cultured for a further 24 h and multicellular embryos with up to 20% of fragments were frozen slowly with 1.5 M dimethylsulphoxide on day 3 post-insemination (162 embryos in CC-HMG cycles, 102 embryos in BUS-HMG cycles). Outcome was measured by embryo survival rate, embryo implantation rate and delivery rate in fresh and frozen embryo transfers. Delivery rates were 31.3 and 21.7% per fresh embryo transfer in BUS-HMG and CC-HMG cycles respectively. Fresh embryo implantation rates were significantly higher in collection cycles stimulated with BUS-HMG (17.9%) than in cycles stimulated with CC-HMG (11.3%). Implantation rates were significantly enhanced in embryos transferred in excess of one in cycles leading to pregnancy, perhaps indicative of higher embryo quality in BUS-HMG cycles. Almost all cryopreserved embryos have by now been thawed, so the contribution of frozen embryos to overall pregnancy rates can be evaluated. Overall morphological survival rates of frozen-thawed embryos were similar for 140 embryos from CC-HMG cycles (50%) and 100 embryos from BUS-HMG cycles (46%). The percentage of fully intact embryos was, however, significantly lower in the BUS-HMG group (19%) than in the CC-HMG group (39.5%). Delivery rates were significantly lower following 30 transfers of frozen-thawed embryos from BUS-HMG-stimulated cycles (3.3%) than following 42 transfers of frozen-thawed embryos from CC-HMG cycles (19.1%). Embryo implantation rates were lower for frozen-thawed embryos from BUS-HMG cycles (2.3%) than from CC-HMG cycles (12.7%). Here we demonstrate that ovarian stimulation with the long protocol BUS-HMG instead of the CC-HMG protocol led to higher embryo implantation rates in collection cycles but to lower intact embryo survival rates and to lower embryo implantation rates for frozen sibling embryos. Despite the lower implantation rates with frozen embryos originating from the BUS-HMG protocol, there was no significant difference between total delivery rate per transfer from cycles stimulated with CC-HMG (30.2%) compared with BUS-HMG (33.8%).