Pulmonary oxygen transfer, defined by PaO2/FIO2, and radiologic presence of atelectasis were measured pre-, intra-, and postoperatively to postoperative day 9 in elective cardiac aortocoronary bypass surgical patients, who were randomly allocated either to receive 18 h PEEP while on the ventilator followed by 12 h of nasal continuous positive airway pressure (nasal CPAP) or to be control subjects. The two groups were comparable in age, sex, forced expiratory volume in 1 sec (FEV1), the ratio of FEV1 over forced vital capacity (FVC), time on pump, units of blood transfused, New York Heart Association grading, and cardiac performance indices. The PaO2/FIO2 was significantly (p less than .05) better from half an hour after extubation until 24 h postextubation in the nasal CPAP group, but was decreased for the remainder of the study in both groups. Incidence of atelectasis/consolidation was not different in both groups during the study period. We conclude that nasal CPAP is well tolerated as a treatment of hypoxemia in the immediate postoperative period of aortocoronary bypass patients. CPAP does not change the course of postoperative atelectasis.