The decision to place an infant on extracorporeal membrane oxygenation (ECMO) is based on predictions of expected morbidity and mortality. One unknown factor is the relationship between pre-ECMO pulmonary dysfunction and on barotrauma and post-ECMO pulmonary sequelae. To determine whether placement of infants on extracorporeal membrane oxygenation (ECMO) early is associated with less subsequent pulmonary dysfunction than placing infants on EMCO later, we evaluated pulmonary function in 25 neonates prior to ECMO, when the infants had come off EMCO, and at the time of nursery discharge. Pulmonary resistance (R) and compliance (CL) were determined by a pneumotachograph and esophageal manometry, and functional residual capacity (FRC) was determined by a helium dilution method. Maximal expiratory flow (VmaxFRC) was determined by thoracic compression at the time of discharge. Infants were assigned to an early ECMO group (< 36 hours of age, n = 12), or a late ECMO group (> 36 hours of age, n = 13). When first evaluated, the early group had a higher oxygenation index than the late group (mean value, 63 versus 48), but initial pulmonary function measurements were not different between the two groups. In the early group mean CL increase from 0.20 to 0.36 ml/cmH2O/kg, FRC increased from 7 to 20 ml/kg, and mean R decreased from 107 to 61 cmH2O/L/sec between the initial study and immediately after ECMO. In the late group, only FRC increased from a mean of 8 to 20 ml/kg. CL and FRC increased from post-ECMO to discharge in both groups (mean CL from 0.36 to 0.76 ml/cmH2O/kg in the early group, and from 0.30 to 0.79 in the late group). Mean FRC increased from 20 to 26 ml/kg in the early group, and from 20 to 25 ml/kg in the late group. VmaxFRC was lower in the late than the early group at discharge (mean, 1.14 versus 1.58 L/sec; P < 0.05). While both groups of infants had minimal pulmonary dysfunction at discharge, the infants placed on ECMO early had evidence of slightly less airway dysfunction despite a higher initial oxygenation index than the infants placed on ECMO late.