Non-small-cell lung cancer patients who have unresectable mediastinal disease have a poor prognosis regardless of therapy administered. The various rationales for high-dose radiation therapy with curative intent are reviewed in the context of treatment results reported in the literature. The case for a dose-cure relationship in a clinically practical dose range must be considered unproven. On the other hand, the evidence for dose-response effects suggests that higher dosages could result in improved quality of life. There is a need to further refine reproducible criteria, both anatomic (such as mediastinal involvement) and non-anatomic (such as performance status), to permit more appropriate selection of patients for high-dose treatment. Studies involving less rather than more treatment with appropriate endpoints might be both ethical and appropriate.