Objective: To investigate the effect of serum estradiol (E-2) levels on the day of human chorionic gonadotropin (hCG) administration on the outcome of controlled ovarian hyperstimulation (COH) in both long gonadotropin-releasing hormone (GnRH) agonist and GnRH antagonist protocols. Materials and Methods: This study included 212 in vitro fertilization-embryo transfer (IVF-ET) cycles performed with either long GnRH agonist or GnRH antagonist protocols were classified into three groups according to serum E-2 levels measured on the day of hCG injection: <2,000 pg/ml, 2,000-4,000 pg/ml, and > 4,000 pg/ml. The three groups were compared according to age, number of retrieved oocytes, number of transferred embryos, and pregnancy rates for each of the stimulation protocols. Results: The long and antagonist protocols were performed in 130 and 82 cycles, respectively. The pregnancy rates were 21.5% (28/130) and 23.2% (19/82) in the long- and antagonist-protocol groups, respectively. Serum E-2 levels were measured on the day of hCG administration as <2,000 pg/ml in 65 cycles, 2,000-4,000 pg/ml in 76 cycles, and > 4,000 pg/ml in 71 cycles. The number of retrieved oocytes increased in parallel to serum E-2 levels (p = 0.001). However, there was no significant difference among groups in the pregnancy rates (p = 0.116). Similarly, the number of retrieved oocytes increased in parallel to serum E-2 levels in both of the protocol groups (p value was 0.001 in both long GnRH agonist and antagonist protocols), but there was no correlation between the pregnancy rates and serum E-2 levels (p value of long GnRH agonist protocol was 0.254 and the p value of antagonist group was 0.349). Conclusion: The serum E-2 level on the day of hCG administration does not predict the pregnancy outcome in IVF with either long GnRH agonist or GnRH antagonist protocols.