OBJECTIVES: To evaluate associations between insurance coverage and length of stay, death, and discharge destination of patients hospitalized for non-hemorrhagic stroke. METHODS: A retrospective analysis was conducted of patients hospitalized for < 30 days with a primary diagnosis of nonhemorrhagic stroke (ICD-9 code = 436) identified from a 10% sub-sample (745,099 cases) of the Health care Cost and Utilization Project (HCUP) 2000 database. The database contains hospitalizations from hospitals in 28 U.S. states. Associations between insurance coverage (Medicare, Medicaid or private/HMO) and length of stay, in-hospital death, and discharge destinations were analyzed using ANOVA, logistic and multinomial regression respectively, controlling for age, gender, race, admission type, admission source, comorbidities and socioeconomic status. Statistical analyses of the data were performed using SAS for Windows Version 8.2. An alpha of <0.05 was required for significance. RESULTS: Insurance type was not statistically significant in predicting LOS after adjusting for other risk factors (p = 0.2095). The interactions between insurance coverage and age, and between insurance and admission source were significant. After controlling for risk factors, insurance type had a significant association with in-hospital death rate (p = 0.0335). Medicare cases were significantly more likely to be discharged to home-health care (p = 0.0006), while private/ HMO cases were less likely to be discharged to home-health care (p = 0.0071), even after controlling for risk factors. CONCLUSIONS: Insurance coverage had a significant association with in-hospital death risk and discharge destination in patients hospitalized for non-hemorrhagic stroke. Further study is needed to elucidate the basis for these associations.