To test the hypothesis that respiratory control is altered in cocaine-exposed infants, we evaluated the hypoxic arousal response and the ventilatory response to carbon dioxide (CO2) in 18 term newborn infants prenatally exposed to cocaine and in 10 healthy, term newborn infants within the first week of life. Three infants could not be tested for the hypoxic arousal response because of low baseline oxygen saturation, and data from these infants were excluded from analysis. Twelve hour overnight pneumocardiograms were performed on all infants. Results show that 60% (9/15) of the prenatally cocaine-exposed infants had an abnormal hypoxic arousal response and 87% (13/15) had abnormal hypercarbic ventilatory response. Only 6% (1/15) of the prenatally cocaine-exposed infants demonstrated any abnormality on pneumocardiogram. In contrast, all control infants (10/10) were aroused by the hypoxic challenge and 80% (8/10) had normal ventilatory response to CO2. No abnormalities were found in the assessment of the overnight pneumocardiogram in the control infants. For the cocaine-exposed infants, test abnormalities were not correlated with a concurrent positive urine toxicology for cocaine, suggesting that the injury occurs early in development. These findings support the hypothesis that infants prenatally exposed to cocaine demonstrate abnormalities of respiratory control.