Objective : To evaluate the safety and efficacy of selective fetal reduction (SFR) and compare the outcome of twin pregnancy after SFR in multiple pregnancy induced by assisted reproductive technology (ART) with that of natural twin pregnancy. Methods : From September 1995 to March 2002 in Ajou University Hospital, SFR was performed in 79 patients whose gestational sacs were more than 3. Of these 79 patients, 47 patents resulted in twin pregnancy after SFR. SFR was performed using transvaginal intracardiac KCl injection at gestational age of $6{sim}9$ weeks. Control group was composed of 264 patients with natural twin pregnancy, who delivered after intrauterine pregnancy at 24 weeks, from June 1994 through December 2002. We compared Obstetric and perinatal outcomes between SFR group and natural twin group. Results: Among 47 patients with twin pregnancy after SFR, 2 spontaneous abortion were occurred at intrauterine pregnancy at 8 and 19 weeks. Obstetrical and perinatal outcomes were available in 43 patients. Single intrauterine fetal death was occurred in 1 of 43 (2.3%) patients in SFR group. Incidence of preterm labor, premature rupture of membrane, preeclampsia and placenta previa were similar, but gestational diabetes mellitus (GDM) was occurred more frequently in SFR group (3 (7.0%) vs 4 (1.5%), p=0.02). Mean gestational age, mean birth weight, incidence of discordancy, use of intubation and ventilation, incidence of fetal anomaly, low (<7) Apgar score and intrauterine growth restriction were similar in both groups. Conclusion: Twin pregnancy after SFR has the increased incidence for GDM but other obstetric and perinatal outcome was similar compared with natural twin pregnancy. So SFR is a safe and effective procedure, so we suggest SFR is needed in multifetal pregnancy more than triplet.