A trial of partial liquid ventilation (PLV) with sterile perflubron(LiquiVent©; Alliance Pharmaceutical, San Diego CA.) is being conducted for infants with severe respiratory failure requiring ECMO support. Infants are enrolled, following informed parental consent, if they have been on ECMO for at least 48 hours, and are not responding to therapy. The lungs are filled with an FRC of perflubron and the infant is maintained on ECMO support and conventional ventilation. Evaporative losses of perflubron are replaced. PLV is continued for up to 96 hours if improvement is noted. To assess the effect of PLV on lung function, compliance (CL, ml/cmH2O/kg,) is determined by airway manometry and pneumotachography. Six infants (ARDS = 4, CDH = 2) have been studied (Mean ± SE age 3 ± 1.6 months) following prolonged periods of ECMO without improvement (mean ± SE 13± 3.3 days on ECMO, range 5 to 21 days). The infants tolerated PLV well, with volume recruitment observed on chest radiograph, and debris removal occurring on suctioning. The total amount of perflubron instilled varied with each infant (mean ± SE 64 ± 19 ml/kg, range 16 to 125 ml/kg). Compared to pre-PLV, all infants demonstrated improvements in Cl during PLV, ranging from 24 to 344% increase (Mean ± SE 186 ± 51%). Four of 6 infants weaned from ECMO to conventional ventilation for at least 3 days. One infant with CDH died of respiratory failure and one infant died from generalized Herpes soon after ECMO was withdrawn. Two infants died at 4 and 68 days post-PLV, and two are long term survivors. None of the deaths were related to PLV. These data demonstrate that PLV recruits lung volume and improves CL in critically ill infants on ECMO. The impact of PLV on long term outcome in these critically ill infants, however, will need to be evaluated. Supported in part by NIH MO1RR00240.