Oxygen arterial saturation (Sao2) was measured with a pulse oximeter in 209 patients after elective surgery. Measurements were made upon arrival in the recovery room (RR) and 1 hr later. On each occasion, the patients randomly and alternately breathed—for 10 min at a time—room air or 35% O2. Factors that might influence the incidence of postoperative hypoxemia were analyzed. After breathing room air for 10 min after arrival in the RR, the mean Sao2 was 90.7 ± 3.9% (±SD). Twenty min after and 1 hr after arrival in the RR, mean Sao2 increased significantly to 92.4 ± 3.5% (P < 0.002) and 93.2 ± 3.0% (P < 0.001), respectively. Postoperative hypoxemia (Sao2 ± 90%) after breathing room air for 10 min at 10 min, 20 min, and 1 hr after arrival in the RR occurred in 43.8%, 26.9%, and 16.9% of the patients, respectively. Breathing 35% O2 for 10 min 10 min after arrival in the RR, as well as 20 min and 1 hr after arrival, significantly increased Sao2 above the Sao2 level after breathing room air by 5.7% (P < 0.001), 4.3% (P < 0.001), and 4.0% (P < 0.001), respectively. A significant multiple correlation was found between low Sao2 levels while breathing room air on arrival in the RR and fentanyl dose, age, and concentration of halothane used intraoperatively (R = 0.46; P < 0.001). The data demonstrate that, in the first minutes in the RR, there is a high incidence of hypoxemia, that administration of 35% O2 relieves arterial desaturation, and that the type of anesthetic and age are the main factors influencing early postoperative hypoxemia.