Although there are effective medical and surgical modalities available for the treatment of patent ductus arteriosus (PDA), it remains a common problem for preterm infants, and its associated morbidities influence the outcomes of these infants. We questioned whether there was a difference in outcomes between preterm infants with PDA who were initially treated with either surgical ligation or indomethacin. We retrospectively reviewed the medical charts of 153 preterm infants born at less than 35 weeks gestation between December 1991 and December 1994. All infants diagnosed with PDA were treated either with ligation alone (n=43), indomethacin alone (n=74), or with indomethacin initially and subsequent ligation (n=36). To control for differences in severity of illness among infants, the Score for Neonatal Acute Physiology (SNAP) was determined for each infant within the first 12 hours of life and again at the time of PDA diagnosis. Using length of hospital stay as a measurement of outcome, we performed a multiple regression analysis controlling for SNAP scores, as well as one and five minute apgar scores, multiple birth, administration of antenatal betamethasone, the presence of hyaline membrane disease, and PDA size. We found that on average, infants treated with indomethacin alone had a 10 day increase in length of hospital stay as compared to infants who were ligated initially (p=.10). Infants treated initially with indomethacin who subsequently required ligation had a 15 day increase in length of hospital stay as compared to the group ligated initially (p=.05). These results indicate that the initial mode of treatment chosen for infants with PDA influences their clinical course.