OBJECTIVES: This study compares the patient characteristics, insurance coverage, disease prevalence, and utilization patterns of inpatient and emergency department (ED) services between homeless and non-homeless people using a new data source that has not been employed in existing studies.   METHODS: A retrospective data analysis was conducted to compare differences in patient characteristics, insurance coverage, disease prevalence, and utilization patterns between homeless and non-homeless individuals who visited a hospital or hospital-based emergency department. For each service type, separate rates were created for homeless and non-homeless populations and then compared. The Healthcare Cost and Utilization Project (HCUP) State Inpatient Databases (SID) and State Emergency Department Databases (SEDD) for 2008 were used in the analysis. The SID employed in this study include 15.9 million inpatient hospital discharges from community hospitals in ten states – Arizona, California, Colorado, Florida, Georgia, Massachusetts, Missouri, New York, Pennsylvania and Wisconsin.  177,056 of these discharges were homeless patients.  The SEDD employed in this study include 23.7 million visits to ED, where patients were treated and released, from seven states (i.e., Arizona, Florida, Georgia, Massachusetts, Missouri, New York and Wisconsin).  49,595 of these visits were homeless patients. RESULTS: The uninsured homeless (non-homeless) patients were accountable for 28.1% (4.6%) of inpatient admissions and 42.85 (21%) of ED visits. Medicaid covered 48.2% of all inpatient discharges and 34.7% of all ED visits by homeless patients.  73.7% (50.6%) of homeless (non-homeless) inpatient admissions came through EDs.  Homeless patients with mental disorders accounted for 22.4% of all homeless inpatient discharges and 49% of all homeless ED visits. A majority of the mental disorder diagnosis in both settings of care were alcohol-related disorders, mood disorders, and schizophrenia. CONCLUSIONS: The profile of homeless and non-homeless patients differed significantly by insurance status and by race in both the inpatient and emergency department settings.