For a period of 35 months, 50 patients presenting with a total of 61 peripheral pulmonary nodules were operated on under videothoracoscopy. As a matter of principle none of these nodules were marked radiologically pre-operatively. All the scanners were reviewed retrospectively by a radiologist and a thoracic surgeon without knowing the results of the thoracoscopic intervention: 23 of these patients on the evidence would have quite obviously required preoperative marking (group I), and 27 would have been presented for direct thoracoscopy (group II). In group I there was only one group of nodules which could not be localised and by necessity, a thoracotomy was required. In group II, two nodules could only be localised thanks to a mino-thoracotomy. The level of failure was between 4 and 7%, and was identical to that found in the literature for different techniques of pre-operative radiological marking: these techniques were often complicated by a pneumothorax and intrapulmonary haemorrhage. These techniques for marking are used extensively. Prospective studies based on precise and complete criteria should enable better definition of rare cases which might benefit.