The current American Joint Committee on Cancer (AJCC) staging system for bronchogenic carcinoma, which divides stage III M0 cases into stages IIIA and IIIB, is based on the observation that selected patients with IIIA disease (T3 or N2) can undergo complete surgical resection, in distinction to IIIB patients (T4 or N3). To understand the value of this system when applied to clinically staged (CS) patients treated with a standard nonoperative approach, the records of patients with squamous cell, large-cell, and adenocarcinoma of the lung treated with radiation therapy (RT) at the Fox Chase Cancer Center from 1978 to 1987 were reviewed. Three hundred sixteen patients were identified as having CS III M0 disease treated with single daily fraction RT without chemotherapy or sensitizers. Of these, the distinction between IIIA (166) and IIIB (140) could be made for 306 patients. The median survival time (MST) for all CS III patients was 9.6 months, and the 2-year survival was 17%. No difference was observed in MST between CS IIIA and IIIB patients (9.4 v 9.8 months, P = .78), in 2-year survival (17% v 18%), or in rate of first failure within the RT field (43% v 44%). MSTs for the 157 CS IIIA and IIIB patients with less than 5% weight loss and Zubrod performance status (PS) 0 to 1 were 13.0 and 15.8 months (P = .29), respectively. This lack of difference in outcome for CS IIIA and IIIB patients receiving RT has important implications in the design and stratification of future nonoperative trials for stage III lung cancer.