Of 95 infants treated with the synthetic surfactant, Exosurf(R), under a Treatment Investigational New Drug protocol, 17 received one dose, 40 received two, and 38 received three doses. Seventy-six (80%) of the infants were treated by rescue protocol. We retrospectively reviewed the clinical course of the 67 surviving rescue infants. We found that, compared to one- and two-dose infants, those treated with three doses of Exosurf(R) were more premature, smaller, required a longer ventilator course, and had more frequent complications, including patent ductus arteriosus (PDA), intraventricular hemorrhage, nosocomial pneumonia, and apnea. They required higher oxygen concentrations starting 8 hr after their first dose and higher mean airway pressure (MAP) from the time of their second dose. These trends continued during all subsequent time points, as compared to infants treated with two doses. The third dose was administered an average of 17 hr after the second, resulting in little change of MAP, but some reduction in oxygen requirements. By 24 hr after the last dose, only 4% of three-dose infants were extubated compared with 30% of the two-dose and 71% of one-dose infants. In conclusion, repeated administration of Exosurf(R) is not equally effective in every treated infant with respiratory distress syndrome (RDS) and complications of prematurity may affect or accompany poor response. We speculate that (1) the group of infants treated with three doses of Exosurf(R) had a greater degree of surfactant deficiency than the other groups, (2) PDA may account for this refractory response to surfactant, and (3) the three-dose group might have responded better to earlier treatment and/or more surfactant doses, or to earlier PDA closure.