Postoperative hypoxemia has been well documented in adults but not in infants and children, although they are potentially more susceptible to airway closure and to disturbances in pulmonary gas exchange. In a prospective study, we measured arterial oxygen saturation (SaO2) with a pulse oximeter in 97 ASA class I infants and children breathing room air before and after general anesthesia for superficial surgical procedures. Mean preoperative SaO2 was 97.6 +/- 0.15% (SEM). On arrival in the recovery room after anesthesia mean SaO2 in room air had decreased significantly (P less than 0.01) to 93.0 +/- 0.49% (range 100-71%), corresponding to calculated arterial oxygen tension (PaO2) of about 66 mm Hg. The second reading, 5-15 min later, also showed a statistically significant (P less than 0.01) decrease in SaO2 (94.1 +/- 0.35%). There was no statistical difference in SaO2 between patients who received inhalation anesthesia alone and those who were given narcotics. There was also no correlation between postoperative reduction in SaO2 and duration of anesthesia or patient age. Of 67 patients who were asleep on arrival in the recovery room, 47 who remained asleep at the second SaO2 reading had an average increase in SaO2 of less than 1%. In contrast, in those patients who awoke, average SaO2 increased more than 4% during a similar time period--a difference that was statistically significant (P less than 0.02).