The minute ventilation to carbon dioxide (CO 2 ) production ratio (V' E /V' CO2 slope) was recently identified as a mortality predictor after lung surgery, but the effect of the resection extent was not taken into account. The aim of this study was to investigate the role of V' E /V' CO2 slope as preoperative mortality predictor depending on the type of surgery performed. Retrospective analysis was performed on 263 consecutive patients evaluated before surgery for lung cancer. Death within 30 days and serious respiratory complications were considered. Univariate and multivariate regression analyses were used to identify independent predictors of death. Lobectomy or bilobectomy were performed in 186 patients with 29/186 (15.6%) serious pulmonary complications and 6/186 (3.2%) deaths. Pneumonectomy was performed in 77 patients with 14/77 (18.2%) serious complications and 5/77 (6.5%) deaths. Considering the whole group, the peak oxygen consumption (V' O2peak, L/min; z=-2.66, p<0.008, OR 0.007) and V' E /V' CO2 slope (z=2.80, p<0.005, OR 1.14) were independent predictors of mortality whereas in pneumonectomies V' E /V' CO2 slope (z=2.34, p<0.02, OR 1.22) was the only independent predictor of mortality. High V' E /V' CO2 slope, age and low V' O2peak are predictors of death and severe complications after lung surgery. Before larger resections as pneumonectomies an increased V' E /V' CO2 slope represents the best mortality predictor.