OBJECTIVE: The purpose of this study was to test the reproducibility and clinical feasibility of three functional performance measures and five single-joint or multijoint muscle power measures. DESIGN: Twenty patients with a mean age of 68.7 ± 7.2 yrs with severe hip or knee osteoarthritis were assessed for test-retest reliability and agreement on two occasions 1 wk apart. The outcomes were maximal single-joint muscle power (hip extension/abduction and knee extension/flexion), maximal muscle power during multijoint leg extension press, and functional performance measures (20-m walk, five-time repeated chair stands, and repeated unilateral knee bending). RESULTS: For single-joint and multijoint maximal peak power and functional performance measures, we demonstrated poor (CVws, approximately 25%, single-joint hip extension) and moderate (CVws, approximately 15%, multijoint leg extension press, single-joint knee extension, chair stands, and knee bending) to good (CVws, 0.81, single-joint knee extension, knee flexion, hip abduction, 20-m walk, and chair stands) reliability. CONCLUSIONS: Isolated muscle power over the hip and knee can safely be evaluated with poor to good agreement and good to excellent reliability in patients with advanced hip or knee osteoarthritis. Functional performance and muscle power may be assessed concurrently.