Suspected benefits of a prospective payment system (PPS) in which hospitals are paid by diagnosis-related groups (DRGs) are that hospital lengths of stay and costs may be reduced. The US Department of Veterans Affairs is one of the first agencies to adopt PPS funding for rehabilitation; this early adoption of PPS provides a unique opportunity to test for both beneficial and adverse outcomes. This study compared hospital stay, readmission rate, and incidence of nursing home placement before and after introduction of PPS on a 22-bed rehabilitation service. Hospital stay decreased from 29.3 days to 26.4 days, but 64% more patients were discharged to nursing homes. Findings suggest that PPS may overlook home care in favor of placement, which neutralizes the cost benefits of the proposed reimbursement system. Further research on the effects of PPS is needed to determine (1) impact on clinical aspects of rehabilitation and (2) whether other funding mechanisms are more appropriate.