STUDY OBJECTIVE: To determine the risk of thoracic and major abdominal surgery in patients with chronic obstructive pulmonary disease (COPD). DESIGN: Retrospective cohort study with controls. SETTING: A 692-bed teaching hospital. PATIENTS: A cohort of 26 patients with severe COPD (FEV1 < 50 percent predicted) undergoing thoracic and major abdominal surgery was matched by age and type of operation to 52 patients with mild-moderate COPD and 52 patients with no COPD. MEASUREMENTS AND RESULTS: The 26 patients with severe COPD had rates of cardiac, vascular, and minor pulmonary complications similar to patients with mild-moderate COPD and without COPD, but experienced higher rates of serious pulmonary complications (23 percent vs 10 percent vs 4 percent, p = 0.03) and death (19 percent vs 4 percent vs 2 percent, p = 0.02). All deaths and instances of ventilatory failure in the patients with severe COPD occurred in the subset undergoing coronary artery bypass surgery. Logistic regression revealed that increased age, higher American Society of Anesthesiologists class, an abnormal chest radiograph, and perioperative bronchodilator administration were associated with higher cardiac or serious pulmonary complication rates. Spirometry was not an independent predictor of postoperative complications. CONCLUSIONS: Clinical variables appear better than preoperative spirometry in predicting postoperative cardiopulmonary complications. The utility of preoperative spirometry as well as the benefits of perioperative bronchodilators in patients in stable condition remain to be determined.