Patients who are considered for major pulmonary resection are normally evaluated by spirometry and clinical assessment. Despite this, the morbidity and mortality rates are high after these operations. We retrospectively reviewed results of lung resection performed during a period of 7.5 years in 237 patients to identify other important predictors of morbidity and mortality. There were 144 male and 93 female patients with a mean age of 59.4 +/- 11.4 years. The indication for operation was lung cancer in 199 (76 stage I, 34 stage II, 89 stage IIIA-B), benign disease in 34, and metastatic disease from other primary tumors in four. Lobectomy or bilobectomy was performed in 164 patients and pneumonectomy in 73. Data on 38 preoperative and operative risk factors were correlated with information on 24 postoperative events grouped into four major categories: death, pulmonary complications, cardiovascular complications, and other problems. Logistic regression analysis and chi 2 analysis were used to identify the relationship of the preoperative risk factors to the grouped postoperative complications. The diffusing capacity of the lung for carbon monoxide was the most important predictor of mortality (p less 0.01) and was the sole predictor of postoperative pulmonary complications (p less than 0.005). This diffusing capacity can reveal the existence of emphysematous changes in the lung, even when spirometric values are acceptable, and it usually should be a part of the evaluation of patients being considered for pulmonary resection.