We evaluated the diagnostic value of pulse oximetry during sleep in pediatric obstructive sleep apnea syndrome (OSAS) caused by adenoid-tonsil hypertrophy. Subjects were 22 healthy children free of symptoms such as snoring, sleep apnea and oral breathing and 163 children suspected of OSAS with snoring or sleep apnea. Subjects were measured for percutaneous oxygen saturation (SpO2) during sleep. Of those with suspected OSAS, 69 underwent adenotonsillectomy and were measured for SpO2, both pre- and postoperatively, then pre- and postoperative measurements were compared. After measurement, we analyzed three parameters: lowest saturation (LSpO2), the desaturation index, and total desaturation duration under 95% (TDD95). Few abnormal findings were seen in healthy children. We calculated the mean and standard deviation (SD) of each parameter and set borderlines of mean-2SD for LSpO2 and mean + 2SD for ODI and TDD95. With these borderlines, 105 children for LSpO2, 75 for ODI and 76 for TDD 95 were judged to be normal among the 163 with suspected OSAS. Histograms showed that the mode of each parameter was situated near the borderline. Comparison between pre- and postoperative measurements showed that the effect of the surgery strongly correlated with preoperative measurement in patients undergoing surgery. Assuming that a patient with postoperative improvement is positive, we calculated sensitivity and specificity for each borderline measurement. We found that if success is 100%, the borderline should be 87% for LSpO2, 3.5 for ODI, and 30.0 for TDD95. If success exceeds 90%, the borderline should be 90% for LSpO2, 2.0 for ODI, and 7.0 for TDD95. We therefore conclude that measurement of SpO2 during sleep is useful in screening for pediatric OSAS.