We undertook to develop a tool based on the FIM™ instrument to predict the number of nursing hours required to care for stroke patients in an acute inpatient rehabilitation program. The initial study to evaluate the feasibility of using the FIM instrument revealed that the total FIM score had a strong inverse relation to the level of care indicated by the Patient Care Index (PCI) at days 1, 5, 7, 10, 15, and 20 of rehabilitation (rs = -.76 to -.87). The results warranted continued investigation of the FIM instrument as a guide for nurse staffing decisions. Based on data from the initial study, five categories of FIM score ranges were designated that demonstrated the most accuracy of placing patients at the correct level of care. Special care considerations unique to institutional settings were identified and incorporated into the tool's final format, as were the calculations to determine the amount of assistance needed. The study reported here was undertaken to evaluate the level of care indicated by the adapted tool, compared with that of the PCI, in a sample of 67 stroke admissions. Spearman correlations revealed a moderate relationship (rs = .49 to .54) between the amount of care determined by the Patient Acuity and Staffing tool and through the PCI at the first, second, and third team meetings. We conclude that the system is an effective, efficient guide for scheduling nurse staffing on the stroke rehabilitation unit.