This study was performed to observe the changes of oxygen saturation after tracheal extubation, which depends on the following tracheal extubation methods of the group 1. 2A. 2B, has done. One hundred twenty-five healthy patients that had been selected out of 1~15 year-old at random were divided as follows; Group 1; extubation performed in being awake (n=49) Group 2; anesthetized extubation (n=76) 2A: extubation done in 5 min after discontinuing N2O (n=38) 2B: extubation under anesthesia (GOE or GOF) (n=38) The grouping, which has at random been assigned to the patients, was done just before the end of operation. Oxygen saturation was measured continuously by pulse-oximeter (Minolta Pulsox TS-7, Japan) and was recorded in the process of operation, immediately after extubation and at 1.2.4.6.8.10.20.30 min after extubation while they were spontaneously breathng room air. In case oxygen saturation were less than 90%, supplementary oxygen was administered to them. The frequency of hypoxemia was higher in Group 2 (19.7%) than in Group 1 (10.2%). In group 2B, 2 patients developed severe hypoxemia and 1 patient developed PVC immediately after extubation. Changes of oxygen saturation were as follows; Oxygen saturation in Group 2B was higher than that in Group 1 at each 2 min and 4 min and was also higher than that in Group 2A at 4 min after extubation. As a result, Group 1 is the safest extubation method because of its low risk of hypoxemia. If anestltized extubation must be needed, Group 2A would be preferable to Group 2B method because the frequency of hypoxemia was higher in Group 2B than in Group 2A.In addition, It is suggested that monitoring oxygen saturation continuously by the patients should be safe.