This study developed a case-mix patient classification system for long-term care. Patient assessment and resource consumption data, collected for 1,151 patients within 23 hospital-based and freestanding long-term care facilities in California, were used to develop a patient classification system made up of 13 homogeneous patient groups. The 13 groups, formed on the basis of nine patient assessment variables, explained 68.5% of the overall variance in resource use. The relatively high variance reduction achieved in the creation of a limited number of groups demonstrates the feasibility of developing measures of case mix based on patient resource use for long-term care. Case-mix classification may provide a useful tool for long-term care reimbursement reform.