Background: Chronic obstructive pulmonary disease (COPD) has been considered a risk factor for postoperative respiratory failure after general anesthesia. However, the association between COPD severity and postoperative respiratory failure among COPD patients is unknown. Our aim was to compare the prevalence of postoperative respiratory failure in COPD patients according to disease severity after general anesthesia.Methods: We retrospectively reviewed COPD patients undergoing spinal surgery with general anesthesia at our clinical center between January 2016 and January 2021. These subjects were divided into four groups (group I = mild COPD, group II = moderate COPD, group III = severe COPD, and group IV = very severe COPD) according to their preoperative lung function. The primary endpoint was a respiratory failure 1 week after surgery. The diagnosis of respiratory failure was made with the presence of one or more of the following criteria: prolonged ventilator dependence, unplanned postoperative intubation, and partial pressure of arterial oxygen (PaO2) <= 50 mmHg while the patient was breathing ambient air in the hospital. The extubation time, perioperative PaO2 and partial pressure of arterial carbon dioxide (PaCO2), postoperative lung infection, and length of hospitalization were also compared.Results: A total of 120 patients who underwent spinal surgery with general anesthesia were included in this retrospective study. Postoperative respiratory failure occurred in 0 (0.0%) patient in group I, 1 (1.5%) patient in group II, 1 (2.5%) patient in group III, and 1 (14.5%) patient in group IV 1 week after surgery (p = 0.219). The duration of anesthesia was 243.3 +/- 104.3 min in group I, 235.5 +/- 78.8 min in group II, 196.0 +/- 66.3 min in group III, and 173.1 +/- 63.7 min in group IV (p < 0.001). Preoperative PaO2, PaCO2, intraoperative oxygenation index [a ratio of PaO2 to fraction of inspired oxygen (FiO(2))], and postoperative PaO2 were significantly different among the four groups (p < 0.001, 0.001, 0.046, <0.001, respectively). No significant differences among the four groups were seen in extubation time, pulmonary infection, or hospital stay (p = 0.174, 0.843, 0.253, respectively). The univariate analysis revealed that higher preoperative PaO2 was associated with a lower rate of postoperative respiratory failure (OR 0.83; 95% CI, 0.72 to 0.95; p = 0.007).Conclusion: The severity of COPD as assessed with GOLD classification was not associated with the development of postoperative respiratory failure. However, lower preoperative PaO2 was associated with greater odds of postoperative respiratory failure in COPD patients.