OBJECTIVE To investigate the therapeutic effect of functional exercise augmented by programmable implanted microstimulators on arm and hand function. DESIGN Before and after study. SETTING Implantation was performed in a neurosurgery unit, systems were programmed, and tests were conducted in a university laboratory and subjects exercised at home. PARTICIPANTS Hemiparetic subjects (N=7) with reduced upper-limb function who were at least 12 months poststroke were recruited from the community. No subjects withdrew. INTERVENTION Microstimulators were implanted into the arms and forearms to activate elbow, wrist, and finger extension, and thumb abduction. After training and programming of the system, subjects underwent 12 weeks of functional home-based exercise with stimulation. MAIN OUTCOME MEASURES The primary functional measure was the Action Research Arm Test (ARAT). Impairment measures included upper-limb Fugl-Meyer Assessment (FMA) and tests of motor control (tracking index), spasticity (electromyography stretch index) strength, and active range of motion (AROM). The assessor was not blinded, but scores were validated by an independent blinded observer. RESULTS All subjects were able to perform functional activities at home by using the system. Compliance was excellent, and there were no serious adverse events. Statistically significant improvements were measured (P<.05) in the tracking index (57.3 degrees(2)+/-48.65 degrees(2)), FMA score (6.3+/-3.59), wrist-extensor strength (5.5+/-4.37 N), and wrist AROM (19.3 degrees +/-18.96 degrees). The mean improvement in ARAT score +/- SD of 4.9+/-7.89 was not statistically significant. CONCLUSIONS This study has shown the feasibility of a programmable implanted microstimulator system used at home to perform functional exercises and a reduction in impairment after 12 weeks.