Polygraphic monitoring studies were performed on more than 150 older preterm infants (postconceptional ages of 36 weeks or more) and full-term neonates to evaluate unexplained or persistent apnea. During polygraphic monitoring, 16 infants were observed to have hypoxemia associated with feedings. The feeding hypoxemia was accompanied by irregular respiratory effort and preceded any associated bradycardia. A comparison group of eight infants with similar gestational and postconceptional ages, but without feeding hypoxemia, was selected retrospectively from other infants referred for evaluation of persistent or unexplained apnea. The group with feeding hypoxemia showed evidence of CNS compromise as manifested by significant elevations of the maximum end-tidal carbon dioxide pressure during sleep and abnormal computed tomograms (7/11 v 0/5 in the comparison group). There was no relationship between feeding hypoxemia and sleep apnea or gastroesophageal reflux. Clinical follow-up showed that the feeding hypoxemia resolved with maturation.