Airway obstruction plays an important role in the pathogenesis of apnea in premature infants who have not previously undergone anesthesia. To determine the role of airway obstruction in postoperative apnea, we studied 74 former premature infants by integrated recordings of nasal airflow, pneumocardiography, and pulse oximetry during the initial 2 h of recovery from inhalational anesthesia. Apnea (greater than 6 s) was classified as central, obstructive, or mixed, wherein mixed apnea consisted of central and obstructive apnea within the same apneic episode. Postoperative apnea was observed in 23 infants, ranging in age from 31-48 weeks postconception: 12 had inguinal herniorrhaphy (hernia group) and 11 had other procedures (other group). Of the 268 apneic episodes in the hernia group, 73% were central, 6% obstructive, and 21% mixed. Infants in the other group had 505 apneic episodes, with a distribution nearly identical to that in the hernia group. Central and mixed apnea occurred in all infants experiencing apnea, except in 1 infant, who had only central apnea, whereas obstructive apnea occurred in only one third of the apneic infants. Arterial hemoglobin desaturation was significantly more frequent at the end of mixed and obstructive apnea than after central apnea (P less than 0.01). In both groups, arterial hemoglobin O2 saturation (SpO2) decreased to less than 80% in approximately 35% of mixed and obstructive apneic episodes, compared to approximately 5% of central apneic episodes. SpO2 remained greater than or equal to 90% in over 80% of central apneic episodes, compared to 40% of mixed and obstructive apneic episodes.(ABSTRACT TRUNCATED AT 250 WORDS)