ABSTRACT. Purpose: To investigate improvement in various impairments by exercise interventions in patients with knee osteoarthritis (OA). Methods: We collected data on randomized controlled trials (RCTs) comparing the effects of exercise intervention with those of either nonintervention or psychoeducational intervention in patients with knee OA. Data on pain, stiffness, muscle strength, range of motion, fl exibility, maximal oxygen uptake, and position sense were synthesized. The Grading of Recommendations Assessment, Development, and Evaluation system was used to determine the quality of the evidence. Results: Thirty-three RCTs involving 3,192 participants were identifi ed. Meta-analysis provided highquality evidence that exercise intervention improves maximal oxygen uptake, and moderate-quality evidence that exercise intervention also improves pain, stiffness, knee extensor and fl exor muscle strength, and position sense. The evidence that exercise intervention improves knee extension and fl exion range of motion was deemed as undetermined-quality. Conclusion: In patients with knee OA, improvement in pain, stiffness, muscle strength, maximal oxygen uptake, and position sense with the use of exercise intervention can be expected. Although the quality of evidence of the effect of exercise intervention on range of motion was inconclusive, exercise intervention should be recommended for patients with knee OA to improve various impairments.