PURPOSE: Previous studies have linked the use of positron emission tomography (PET) with improved outcomes among patients with non-small-cell lung cancer (NSCLC). However, this association may be confounded by PET-induced stage migration and selection bias. We examined the association between PET use and overall survival among Medicare beneficiaries with NSCLC. PATIENTS AND METHODS: Retrospective analysis of Surveillance, Epidemiology, and End Results (SEER) -Medicare data was used to characterize changes in overall survival, stage-specific survival, and stage distribution among Medicare beneficiaries with NSCLC between 1998 and 2003. RESULTS: A total of 97,007 patients with NSCLC diagnosed between 1998 and 2003 met the study criteria. Two-year and 4-year survival remained unchanged, despite widespread adoption of PET. The proportion of patients staged with advanced disease increased from 44% to 50%. Upstaging of disease was accompanied by stage-specific improved survival, with 2-year survival of stage IV disease increasing from 8% to 11% between 1998 and 2003. PET was more likely to be administered to patients with less advanced disease (stages I through IIIA) and greater overall survival. CONCLUSION: Overall survival among Medicare beneficiaries with NSCLC was unchanged between 1998 and 2003, despite widespread adoption of PET. The association between PET use and increased survival likely reflects an artifact of selection bias and consequent stage migration.