A prospective double-blind study was designed to assess the efficacy of antibiotic prophylaxis in lung surgery. It included 114 patients undergoing lung surgery for primary or secondary malignant tumours randomly assigned to two groups. Group A patients (n = 59) were given cefamandole intravenously every four hours, three times, starting from induction of anaesthesia. The dose was determined according to the patient's weight: 1.5 g for patients weighing less than 60 kg, 2.5 g for those weighing between 60 and 80 kg, and 3 g for those above 80 kg. Group B patients (n = 55) were given a placebo at the same times. Nineteen other patients were excluded because either the tumour was found to be infected, or the patient had to be mechanically ventilated postoperatively, or an exploratory thoracotomy only was carried out, or they were allergic to beta-lactam antibiotics. The efficacy of antibiotic prophylaxis was assessed by recording the incidence of postoperative infections, the length of the patient's stay in hospital, and the need to use an antibiotic treatment. Patients, their sputum and wound were examined every day, and their temperature recorded. The white blood cell count and chest X-ray was carried out every day for the first week. All the drain and catheter tips were cultured, as well as sputum and blood (every three days). In case of infection, samples were obtained and cultured. Both groups of patients were similar with regard to age, risk factors (smoking habit, diabetes mellitus), and type of surgery (segmentectomy, lobectomy, pneumonectomy). There were 9 postoperative infections in group A, and 22 in group B (p = 0.003).(ABSTRACT TRUNCATED AT 250 WORDS)