OBJECTIVES: There is no study empirically comparing the racial disparity in hospital inpatient cost between homeless and non-homeless patients. This paper compares hospital inpatient costs between homeless and non-homeless patients across race groups. This paper further provides comprehensive descriptive information on patient characteristics, insurance coverage, disease prevalence, and severity of illness for homeless patients and non-homeless patients by race groups. METHODS: We employed propensity score matching to create two cohorts, where inpatient hospital visits between homeless and non-homeless patients were matched based on patients’ demographics and clinical information, and hospital types. These new similar cohorts were then used to compare the cost per hospital visit between homeless and non-homeless patients through multivariate regression analysis. The Healthcare Cost and Utilization Project (HCUP) State Inpatient Databases (SID) for 2009 were used. Data were drawn from 579 hospitals reporting data on homelessness in Arizona, California, Colorado, Florida, Georgia, Massachusetts, Missouri, New York, Pennsylvania, and Wisconsin. SID provide detailed diagnoses, procedures, total charges and patient demographics including homelessness. We linked SID files with American Hospital Association Annual Survey Database and Area Resource File to obtain hospital and area level characteristics. RESULTS: We find significant variations in hospital inpatient costs between homeless and non-homeless patients across race groups. Our risk adjusted results show that homeless patients generally have lower inpatient costs when compared to non-homeless patients. We find that homeless white patients have higher inpatient costs and non-homeless white patients have lower inpatient costs when compared to their homeless non-white and non-homeless non-white counterparts respectively. We also find that the majority of homeless children at hospital inpatient departments are non-white. In addition, about a third of all inpatient hospital visits by white homeless patients are for female individuals. CONCLUSIONS: This study provides comprehensive empirical evidence for racial disparity in hospital inpatient costs between homeless and non-homeless patients.