In two prospective, randomized clinical studies Corynebacterium parvum (C.p.), BCG or 0.9% saline solution were injected intrapleurally once between the 6th and 12th postoperative day in 878 patients with resected, non-oatcell bronchial carcinoma (stage I and II). The rate of complications following C.p.-instillation (chest pain, fever) was relatively low compared to partially serious complications following injection of BCG, especially in patients who underwent pneumonectomy (21.8% empyema compared to 3.2% in the placebo group). After an average follow-up of 4.6 years in the first study, 110 of 207 C.p.-treated patients (53%) were dead compared to 82 of 198 in the placebo group (41%). The difference is statistically significant (P less than 0.02). In the second study, 79 out of 197 BCG-treated patients (40%) were dead after an average follow-up of 2.8 years compared to 71 out of 208 placebo-treated patients (34%). This difference is not statistically significant. From these data it is concluded that postoperative immunotherapy as performed in these studies is not only of no benefit to the patients, but might even be detrimental as a result of an increased posttherapeutic complication rate and a partially diminished expectation of life.