This study of 45,691 patients from 73 hospitals in 22 states shows that Medicaid patients have longer total lengths of stay as well as longer preoperative and postoperative lengths of stay than privately insured patients after adjusting for admission severity of illness and other patient variables. Medicaid patients also had higher ancillary charges and higher total charges. These hospital utilization patterns held in states that paid hospitals per case as well as in states that paid hospitals per diem rates with the exception of ancillary services in the per diem states. A direct test between Medicaid patients under the two payment systems revealed no difference in length of stay, ancillary, and total charges. These findings suggest that service levels to Medicaid patients were not associated with how the hospitals were paid and challenge the assumptions that hospitals can continue to reduce services in response to reimbursement constraints. The implications of these findings for health policy and health services management are discussed.