Preventing hospitalization is one of the major objectives of home health care. Accomplishing this goal is being able to identify patients at risk for hospitalization and intervening appropriately. The current study explored which factors place patients at risk at the start of care and are predictive over the first 60 days of care. Outcomes Assessment Information Set (OASIS), plan of care, medications and medical record information from an urban home health agency were used to build and validate a predictive hospitalization model. The model was developed and tested using a large set of patients (n = 46,366). Patients were classified into seven risk groups from very low to very high. Results revealed that a combination of demographic, financial, clinical and health status factors could accurately predict patients' likelihood for hospitalization and the model agreed with clinical judgments. Examples of how the risk model could be used in practice are provided.