We have been developing exoskeleton robots for assisting the motion of physically weak individuals such as elderly or slightly disabled in daily life. In this paper we propose an EMG-Based control of a three degree of freedom (3-DOF) exoskeleton robot for the forearm pronation/ supination motion, wrist flexion/extension motion and ulnar/ radial deviation. The paper describes the hardware design of the exoskeleton robot and the control method. The skin surface electromyographic (EMG) signals of muscles in forearm of the exoskeleton's user and the hand force/forearm torque are used as input information for the proposed controller. By applying the skin surface EMG signals as main input signals to the controller, automatic control of the robot can be realized without manipulating any other equipment. Fuzzy control method has been applied to realize the natural and flexible motion assist. An experiment has been performed to evaluate the proposed exoskeleton robot.