A prospective controlled randomized study of the use of extracorporeal membrane oxygenation to treat newborns with respiratory failure was carried out using the 'randomized play-the-winner' statistical method. In this method the chance of randomly assigning an infant to one treatment or the other is influenced by the outcome of treatment of each patient of the study. If one treatment is more successful, more patients are randomly assigned to that treatment. A group of 12 infants with birth weight greater than 2 kg met objective criteria for high mortality risk. One patient was randomly assigned to conventional treatment (that patient died); 11 patients were randomly chosen for extracorporeal membrane oxygenation (all survived). Intracerebral hemorrhage occurred in one of 11 surviving children. Extracorporeal membrane oxygenation allows lung rest and improves survival compared to conventional ventilator therapy in newborn infants with severe respiratory failure.