There are no published survival data after chemoradiotherapy (chemoRT) in pathologically documented stage IIIB non-small-cell lung cancer. Studies of radiotherapy (RT) alone or chemotherapy followed by RT yield 5-year survivals less than 10%. The Southwest Oncology Group (SWOG) employed the same concurrent chemoRT induction regimen used in its predecessor trimodality trial to determine the efficacy, safety, and long-term outcome of replacing postinduction surgery with additional chemoRT. Eligible patients for SWOG-9019 had pathologic documentation of T4N0/1, T4N2, or N3 stage IIIB non-small-cell lung cancer. They had pulmonary function adequate to withstand combined-modality therapy, identical to the requirements of the previous trial with postchemoRT surgery. Induction therapy was two cycles of cisplatin plus etoposide (PE) concurrent with once-daily thoracic RT (45 Gy). In the absence of progressive disease, RT was completed to 61 Gy, with two additional cycles of cisplatin plus etoposide. Fifty eligible patients were accrued with tumor-node (TN) substage confirmed on central review: 18, T4N0/1; 12, T4N2; and 20, N3. Grade 4 neutropenia was the most common toxicity (32%). Grade 3/4 esophagitis occurred in 12% and 8%. Median follow-up was 52 months, and overall median survival was 15 months (10 to 22, 95% confidence interval). Three- and 5-year survivals were 17% and 15% (5-year T4N0/1, 17%; T4N2, 13%; and N3, 15%). Feasibility and long-term survival support the application of these results as a standard against which mature outcomes of chemoRT trials with new chemotherapy agents can be compared. These results also justify use of the SWOG-9019 approach as a control arm in ongoing phase III trials.