The purpose of this study was to measure kinesthetic accuracy in healthy older adults by using arm position and motion matching tests. We investigated the effect of task type, joint angle, and matching arm results on kinesthetic accuracy in the upper extremities of 17 healthy right-handed older adults. Blinded subjects were asked to match positions and motions at four reference joint angles: 1) shoulder flexion, 0°–60°; 2) elbow flexion, 90°–135°; 3) wrist extension, 0°–50° in the sagittal plane; and 4) shoulder abduction, 0°–60° in the frontal plane. The absolute difference in angular displacement between the reference and matching arms was calculated to determine kinesthetic accuracy. Results showed that subjects were more accurate at matching motion than position tasks (P=0.03). Shoulder and elbow joints were more sensitive than wrist joints in perceiving passive positions and motions (P<0.05). The effect of the matching arm was found only when matching the joint angles of shoulder abduction and wrist extension (P<0.01). These results are comparable to findings of other studies that used machine-generated kinesthetic stimuli. The manual measurement of kinesthetic accuracy could be effective as a preliminary screening tool for therapists in clinical settings.