BACKGR0UND. A three-arm Phase III randomized trial was performed to compare response rates, time to local or distant progression, and survival for patients with unresectable (Stage IIIA or IIIB) nonsmall cell lung carcinoma treated with standard fractionated thoracic radiotherapy (SFTRT) versus accelerated hyperfractionated thoracic radiotherapy (AHTRT) with or without combination etoposide and cisplatin chemotherapy. METHODS. This trial was initiated in 1992 by the North Central Cancer Treatment Group. Patients with Stage IIIA or IIIB nonsmall cell lung carcinoma were eligible. They were randomly assigned to either SFTRT (6000 centigray [cGy] in 30 fractions) or AHTRT (150 cGy twice daily to a total dose of 6OOO cGy, with a 2-week break after the initial 3000 cGy); the AHTRT was given alone or with concomitant cisplatin (30 mg/m2, Days 1-3 and 28-30) and etoposide (100 mg/m2, Days 1- 3 and 28-30). RESULTS. A total of 110 patients were entered on study. Eleven patients were declared ineligible or off study on the day of study entry. This analysis was confined to the 99 eligible patients. This article reports mature follow-up, because more than 80% of the patients have died. The median follow-up of living patients was 2.5 years. There were suggestions of improvement in the rates of freedom from local recurrence and survival for patients treated with AHTRT (with or without chemotherapy) as opposed to SFTRT (P = 0.06 and P = 0.10, respectively). The improvement in survival associated with AHTRT (with or without chemotherapy) was statistically significant for the subgroup of patients with nonsquamous cell carcinoma after adjustment for other potentially confounding factors (P = 0.02). No differences in freedom from systemic progression or survival were found in a comparison of AHTRT with chemotherapy and AHTRT without chemotherapy. CONCLUSIONS. These results suggest that treatment of Stage IIIA or IIIB nonsmall cell lung carcinoma with AHTRT with or without chemotherapy may improve freedom from local progression and survival as compared with SFTRT, especially for patients with nonsquamous cell carcinoma. The statistical powers to detect the observed differences in median time to local progression and survival were approximately 55% and 35%, respectively. Therefore, further investigation comparing SFTRT with AHTRT is warranted.