Extracorporeal membrane oxygenation will continue to be an appropriate modality of treatment in properly selected infants. In the future, respiratory management and selection criteria for these patients should become standardized and universally accepted. It is conceivable that as we become more comfortable with this treatment we can complete prospective, randomized trials on infants who have less than a 90 percent mortality likelihood and thereby avoid the ethical implications of a study in which the control population has death as an endpoint. It is also conceivable that ECMO will be more benign and induce less morbidity than the barotrauma seen in many of the infants supported by aggressive ventilator management.