Abstract: Objective：To study the pregnancy and delivery rates of repeated IVF-ET cycles，and explore factors associated with pregnant outcome of repeated cycles. Methods：1 746 cycles of oocyte retrieval（128 repeated cycles）were retrospectively analyzed in our reproductive center from 2008.1 to 2009.10. Among the women who had no pregnancy in cycle 1 and cycle 2，there are 68 patients were still not pregnant in cycle 3（group A），and 37 patients got pregnancy in cycle 3（group B）. The factors of cycle 1 and cycle 2 were compared between group A and group B. Results：Clinical pregnancy and live-birth rates for cycle 1，2，3，4 and ≥5 were 35.5% and 34.3%，25.6% and 22.7%，23.4% and 14.0%，15.0% and 12.5%，22.7% and 15.0% respectively（P＜0.001）. The clinical pregnancy rates decreased significantly from cycle 1 to cycle 2（P＜0.05），and the live-birth rates decreased significantly from cycle 1 to cycle 2，3 and 4（P＜0.05），but there was no significant decline in clinical pregnancy and live-birth rates within cycle 2，3，4 and ≥5. It is shown in logistic model that age， primary or secondary infertility， and number of good quality embryo are significant in both cycle 1 and cycle 2. Conclusions：The outcome of the first IVF-ET cycle is better than followed cycles，but the clinical pregnancy and live-birth rates of following IVF-ET cycles were similar. Among patients who did not get pregnancy in cycle 1 and 2，younger，secondary infertility women with more good quality embryos may more likely achieve a desired outcome in cycle 3.

Key words: Fertilization in-vitro, Embryo transfer, Pregnancy rate, Pregnancy outcome, Retrospective studies