We examined 30 newborn infants for respiratory compromise before, during, and after placement in a recommended car seat restraining device. Twelve infants were premature with a history of apnea, eight were premature without known apnea, and 10 were born at term. Both premature groups had significant decreases in oxygen saturation while in the car seat (P less than 0.01) and more frequent desaturation episodes less than 80% (P less than 0.001). Premature infants with a history of apnea had more bradycardia events (P less than 0.05). No normal term infant had any of these problems. In addition, oxygen saturation trended downward from baseline for all premature infants during the recovery interval (P = 0.07). We conclude that currently available car seats may place premature infants at risk for significant hypoxia and ventilatory compromise.