PURPOSE: To report the results of CEBI 140 and 142 trials. These trials were aimed at improving the local control in stage III non-small cell carcinoma with concomitant chemotherapy and radiotherapy in the CEBI 140 trial, and with concomitant chemotherapy and radiotherapy followed by local excision in the CEBI 142 trial. MATERIAL AND METHODS: Thirty-four patients presenting with stage III non-small cell lung carcinoma were included into the CEBI 140 trial from December 1989 to December 1992. Patients were treated with a combination of daily cisplatin (6 mg/m2 per day, 144 mg/m2 in total), vindesine once a week (2.5 mg/m2, 15 mg/m2 in total) and bifractionated radiotherapy (60 Gy/48 fractions/6 weeks) followed by two cycles of cisplatin 120 mg/m2 (at d18 and d45 after completion of radiochemotherapy) and three cycles of vindesine (6 mg/m2 at d24, d31, and d38 after completion of radiochemotherapy). Twenty-eight patients presenting with stage IIIB non-small cell carcinoma-were included into the CEBI 142 trial since January 1993. Patients received a combination of cisplatin (100 mg/m2 at d1 and d24, 200 mg/m2 in total), vinblastine (4 mg/m2 at d1 and d24, 8 mg/m2 in total), 5-fluorouracil in continuous infusion (1,000 mg/m2 from d1 to d3, and from d24 to 26, 6,000 mg/m2 in total) and bifractionated radiotherapy (two series of 21 Gy/14 fractions/9 days, 11 days apart) followed by a new evaluation and surgical excision. RESULTS: In the CEBI 140 trial, all patients received a complete course of radiotherapy, but the dose of cisplatin was decreased in 27% of the cases, and the dose of vindesine in 88%. There were two toxicity-related deaths. Three months after completion of the protocol, there were 50% of complete responders. The overall survival rates at 1, 2 and 3 years were 53, 33, and 11%, respectively, and disease-free survival rates 21 11, and 11%, respectively. In the CEBI 142 trial the immediate tolerance was good. Twenty-one patients (75%) underwent surgical resection. Four tumors could not be resected. Resection was histologically incomplete in one case, and complete in the 16 remaining cases. With a median follow-up of 14 months, ten patients were alive and disease-free. CONCLUSION: Preliminary results of the CEBI 142 trial are encouraging. More patients and longer follow-up are needed for definitive conclusion. It would be of interest to implement a randomized trial comparing the CEBI 142 scheme and classical radiation therapy.