As for other major thoracic operations the conventional 30-day morbidity and mortality marker may underestimate the actual surgical risk of extrapleural pneumonectomy. We retrospectively analysed the prolonged follow-up of 78 patients submitted to extrapleural pneumonectomy for pleural mesothelioma (55), lung cancer with associated carcinomatous (7) or purulent (8) pleuritis, empyema/destroyed lung (4), and mediastinal (2) and chest wall (2) tumours with pleuro-pulmonary involvement. Significant rates of surgery-related major complications (19%) and fatalities (6.6%) additionally occurred beyond 30 days and within 6 months of extrapleural pneumonectomy, making a 66% cumulative (early + late) morbidity rate and an 11.5% cumulative mortality rate, which are respectively 50% and 100% greater than the 30-day rate alone. The leading causes of late morbidity and mortality were respiratory/cardiac sequelae (50%) and broncho-pleural fistulas (30%). Strict preoperative functional selection and proper application of the technical learning curve can reduce the occurrence of the adverse events by anything up to 50% (early mortality: 2.3%). If the results of this novel study of long-term surgical outcomes of extrapleural pneumonectomy were to be confirmed, the preoperative risk/benefit balance of the procedure, mainly when performed for thoracic malignancies, should therefore include the entire spectrum of (early and late) potential surgery-related complications.