In a series of 61 patients in whom pulmonary resection was performed for carcinoma of the bronchus, there was a significant correlation between pre-operatively measured peak expiratory flow rates of less than 70 per cent of the predicted values, and postoperative moratality. A similar correlation was found between pre-operative forced expiratory ratios of less than 60 per cent and postoperative mortality. The correlation between pre-operative vital capacities and postoperative mortality was of less prognostic value. Prevention and recognition of postoperative respiratory insufficiency has been discussed in the light of the pre-operative estimation of respiratory function tests and blood-gas determinations. Prophylactic use of intermittent positve pressure ventilation is recommended as a method of helping to reduce mortality in those patients shown by the pre-opertative tests to be at greatest risk.