Objective: To compare the pregnancy outcomes of transvaginal multifetal pregnancy reduction (MFPR) according to the gestational period when the procedure was performed Methods: From January l995 to February 1998, total 27 patients with multiple pregnancy were included in this study. The patients were grouped to early MFPR group (<8 weeks, n=16) and delayed MFPR group (>8 weeks, n=ll) according to the gestational age that MFPRs were performed. All MFPRs were performed by transvaginal sonography-guided fetal aspiration or mechanical trauma. The complete pregnancy loss rate before 24 weeks of gestation, spontaneous loss of embryo, procedure-related complication, gestational age at delivery, and pregnancy complication were compared between the two groups. Statistical analysis of data was performed using Students t-test and Fishers exact test as appropriate. Statistical significance was defined as p<0.05. Results: There was no significant difference in the complete pregnancy loss rate between the early MFPR group (6.3%) and the delayed MFPR group (27.3%). The incidence of partial spontaneous loss of embryo in the two groups were not differed significantly (6.3% vs. 18.2%). The procedure-related complication of the delayed MFPR group (36.4%) seemed to be higher than that of the early MFPR group (6.3%), however there was no statistical difference (p=0,07). Especially, all 3 patients in whom the MFPR was performed after 10 weeks suffered from the procedure-related complication. The mean gestational age at delivery of the two groups were not differed significantly (36.3+2.8 weeks vs. 37.0+1.3 weeks). There was also no significant difference in the mean birth weights of the two groups (2378.8+563.7 gm vs. 2427.1+436.2 gm). Conclusion: Although there was no statistically significant difference, the early transvaginal MFPR might be a safe and useful method without significant adverse complications compared to the delayed MFPR.