UNLABELLED: Treatment disparities for disabled Medicare beneficiaries with stage I non-small cell lung cancer. OBJECTIVE: To compare initial treatment and survival of nonelderly adults with and without disabilities newly diagnosed with non-small cell lung cancer. DESIGN: Retrospective analyses; population-based cohorts. SETTING: Eleven Surveillance, Epidemiology, and End Results cancer registries. PARTICIPANTS: Persons with disability Medicare entitlement (n=1016) and nondisabled persons (n=8425) ages 21 to 64 years when diagnosed with stage I, pathologically confirmed, first primary non-small cell lung cancer between January 1, 1988, and December 31, 1999. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Initial cancer treatments (surgery, radiotherapy), survival (through December 31, 2001). Multivariable logistic regression and Cox proportional hazards regression estimated adjusted associations of disability status with treatments and survival. RESULTS: Persons with disabilities were much more likely to be male, non-Hispanic black, and not currently married. Although 82.2% of nondisabled persons had surgery, 68.5% of disabled persons received operations. Adjusted relative risks (RRs) of receiving surgery were especially low for persons with respiratory disabilities (adjusted RR=.76; 95% confidence interval [CI], .67-.85), nervous system conditions (adjusted RR=.86; 95% CI, .76-.98), and mental health and/or mental retardation disorders (adjusted RR=.92; 95% CI, .86-.99). Persons with disabilities had significantly higher cancer-specific mortality rates (hazard ratio [HR]=1.37; 95% CI, 1.24-1.51) than persons without disabilities. Observed differences in cancer mortality persisted after adjusting for demographic and tumor characteristics (adjusted relative HR=1.23; 95% CI, 1.10-1.39). Further adjustment for surgery use eliminated statistically significant differences in cancer mortality between persons with and without disabilities across disabling conditions. CONCLUSIONS: Persons with disabilities were much less likely than nondisabled Medicare beneficiaries to receive surgery; statistically significant cancer-specific mortality differences disappeared after accounting for these treatment differences. Future research must explore reasons for these findings and whether survival of disabled Medicare beneficiaries with early-stage, non-small cell lung cancer could improve if surgical treatment disparities were eliminated.