Radiation therapy is widely considered the primary treatment for inoperable "non-small" cell carcinoma of the lung. In clinical investigations, distinction has been infrequent among the histopathologic subtypes of non-small cell carcinoma. Studies have shown significant differences between squamous cell carcinoma and adenocarcinoma/large cell carcinoma; adenocarcinoma/large cell carcinoma has a greater propensity for extrathoracic dissemination, especially to the brain, and it is less curable by resection when regional lymph node metastases are present. No differences have been documented between adenocarcinoma and large cell carcinoma. A retrospective study was undertaken to determine the results of definitive radiation therapy by histopathologic subtype of non-small cell carcinoma of the lung. Between July 1977 and April 1983, 134 patients with non-small cell carcinoma of the lung underwent definitive radiation therapy with curative intent. All patients had performance status scores of 80 to 100 (Karnofsky), and received minimum total doses within the tumor of 60 Gy in 6 to 7 weeks, five fractions per week. The median period of observation was 63 months. Ninety patients had squamous cell carcinoma; 44 had adenocarcinoma/large cell carcinoma. The two groups of patients were comparable in respect to age and Stage; there were significantly more women with adenocarcinoma/large cell carcinoma (27%) than with squamous cell carcinoma (13%). The median survival for patients with squamous cell carcinoma was 11.5 months; the 2 and 4 year survival rates were 21 and 7%, respectively. The median survival for patients with adenocarcinoma/large cell carcinoma was 18 months; 2 and 4 year survival rates were 38 and 23%, respectively. Comparison of the overall survival experience did not show a significant difference between the two cell types (p = .12 using Gehan's generalized Wilcoxon test). However, comparison of the proportion of patients with adenocarcinoma/large cell carcinoma surviving 18 months (50%) was significantly higher (p = .02) than that with squamous cell carcinoma (30%). A small body of data from the literature also suggests a better long-term prognosis for adenocarcinoma/large cell carcinoma. This observation requires confirmation from large trials with histopathologic review. If it is confirmed, there are important implications for therapeutic strategies in future clinical investigations of inoperable carcinoma of the lung.