The aim of the present investigation was to evaluate the predictive value of a symptom-limited exercise test in predicting postoperative morbidity and mortality in patients submitted to pneumonectomy. The study was conducted in 150 patients (mean age, 57.1). Forty-four patients (29.3%) had postoperative complications. Four patients (2.7%) died within one month of the pneumonectomy. Patients with complications had significantly lower VO2max. The incidence of complications in relation to the amount of oxygen consumption showed that with the progressive decrease of oxygen consumption there was a progressive increase in frequency of complications. In particular patients with VO2max < 50% of predicted should be considered at high risk of morbidity and mortality from cardiopulmonary causes. This predictive capacity of VO2max was more evident in patients with a preoperative FEV1 less than 70% of predicted. The present data support the suggestion that exercise testing could be a useful adjunt in the evaluation of postoperative risk for pneumonectomy, especially in patients with obstructive pulmonary disease.