PURPOSE: Indomethacin treatment is successful in about 90% of patent ductus arteriosus (PDA) in premature infants, but in some, repeated administration or surgical closure is required. The object of the present study is to determine the factors affecting the efficacy of indomethacin treatment and to predict the treatment result. METHOD: The 29 preterm neonates, admitted to neonatal intensive care unit of Kyunghee university medical center and Eulji university hospital between September 2002 and April 2006 were diagnosed of PDA and treated with indomethacin. The risk factors that might affect the efficacy of treatment were studied retrospectively. RESULTS: The single-administered group was 19 patients out of 29 (65.5%) and among the repeated- administered group, 5 patients (17.2%) had repeated indomethacin administration and the other 5 patients (17.2%) underwent surgery due to reopening of the duct after repeated medical treatment. In repeated-administered group, the diameter of PDA was significantly larger (3.66+/-0.8 mm vs 2.55+/-0.8 mm, P<0.01), especially when larger than 3.5 mm (sensitivity: 70%, specificity: 89%). Also, the mean postnatal age of the first indomethacin administration was significantly delayed in repeated- administered group (19.7 days vs 12.5 days, P<0.05). There were no significant differences in gestational age (32 wk 5 days vs 30 wk 8 days) and in birth weight (1598.9 g vs 1750.5 g). There were no significant differences in associated morbidities. CONCLUSION: In patients with larger diameter of PDA (>3.5 mm) and older postnatal age (>7 days), the effect of indomethacin was decreased. Therefore in such cases, repeated dose of indomethacin or surgical ligation should be considered earlier.