PURPOSE: Patent ductus arteriosus (PDA) of prematurity is very important disease to study because it causes many cases of perinatal morbidity and its incidence is now increasing. Nowadays indomethacin is the drug of choice for PDA closure, but its use has been limited due to its side effects. Therefore, we compared the effect and side effects of indomethacin according to the infusion method, continuous versus intermittent infusion, to find better an administration method. METHODS: Twenty-five preterm infants who were admitted to Seoul National University Children's Hospital (SNUCH) NICU from March 1995 to August 1996 with a diagnosis of respiratory distress syndrome (RDS) and PDA, were enrolled. They were randomly assigned to intertmittent lV group or continuous lV group. Each group received three intermittent doses or continuous infusion over 36 hours, respectively. We analyzed the perinatal history, time of diagnosis and treatment of PDA, size of PDA, and compared the laboratory parameters, intraventricular hemorrhage (IVH), periventricular echogenecity (PVE), ductal closure and perinatal morbidity before and 48 hours after indomethacin administration between the two groups. RESULTS: Fourteen infants (birth weight 1,149 +/- 373g) were intermittent lV group and eleven infants (birth weight 1,212 +/- 504g) were continuous lV group. There were no significant difference between the groups in perinatal history, pretreatment laboratory parameters, ductal closure, and perinatal morbidity. Patients with IVH of grade 2 or more increased significantly in intermittent lV group (50%) compared to continuous lV group (9%), and PVE progressed significantly in intermittent lV group (64%) compared to continuous lV group (18%). CONCLUSION: Continuous infusion of indomethacin for PDA closure in preterm infants with RDS appears to be as effective as intermittent infusion of indomethacin in closing PDA and have less side effects such as IVH and PVE progress.