30 cases of haemoptysis ‘out of the blue,’ whose x-ray appearances were normal, were classified according to bronchoscopic appearances. In 13 the appearances were normal; 4 had appearances of bronchitis and 5 showed mucopus coming from one bronchus; there was 1 bronchial stricture thought to be tuberculous in nature and 1 inflammatory stricture of unknown aetiology. These 24 cases comprised the non-malignant group of which one died of secondary adenocarcinoma seven months after haemoptysis. The 23 survivors were re-examined after an average period of twenty-three months and, with one exception, have had no further serious trouble. The exception, the only patient shown to have bronchiectasis, had a second haemoptysis. There were 6 patients in the malignant group. 4 had carcinomas for certain and of these 3 died and 1, following removal of his right lower and middle lobes, was well and working twenty-four months after his first haemoptysis. Of the 2 doubtful carcinomas both were alive and no worse than at the time of bronchoscopy, the one nine months and the other twenty-two months later. Though bronchoscopy should never be recommended lightly it is usually worth trying to establish the diagnosis for certain in bronchial carcinoma. It enables the family doctor to manage the patient better after discharge from hospital even when there is no question of advising surgery or x-ray therapy. In the present series this was done in 4 instances and 1 of these patients has, after operative treatment, a reasonable chance of cure. In addition 23 patients were correctly told that there was no carcinoma present and no doubt benefited from this assurance.