The aim of this work was to originally measure mechanical power output of the thumb adductor muscles during fast adduction of the thumb in the horizontal plane. This information will contribute to biomechanical guidelines to help clinicians, sport medicine and rehabilitation specialists in the objective functional evaluation of abnormalities of thumb adductors. Participants performed 20 fast adductions in response to audio signals. Maximum and average angular velocity and angular acceleration were measured. Tangential components of these parameters were then derived. The force of adduction was obtained from the tangential acceleration and the mass of the rotational system. The power was then calculated as the product of the force of adduction and average tangential velocity during the acceleration phase of adduction. All young and untrained males and females were strictly right handed. There was no significant difference in power between dominant and nondominant muscles for either males or females, but males developed significantly more power than females. Because adduction was performed at maximal speed, these data may be explained by the influence of parallel and series elastic elements in the muscle, as well as by influence of fast twitch fibers. Power may be used as a clinical index of the effectiveness of muscle contraction. The similarity of power outputs from dominant and nondominant thumb adductor muscles of right-handers can suggest a classical Bernstein approach. This theoretical approach purports that peripheral factors can distort central commands projected to dominant and nondominant extremities.