A cortical network for voluntary movement control within constraints from neurophysiology and motor psychophysics is developed. A neural controller is proposed to track the desired joint trajectories of a single link controlled by an agonist-antagonist pair of actuators with muscle-like properties. The system is capable of executing voluntary reaching movements, with a typical bell-shaped velocity profile, under unexpected load conditions. The goals of this research work are: (1) to apply the knowledge of human neuro-musculo-skeletal motion control to a biomechanically designed, neural controlled, robotic system, and (2) to demonstrate that such system is able to respond voluntarily similar to human movements on an experimental platform. Through experimental performance results, we shown that neural controller exhibits key kinematic properties of human movements, dynamics compensation, and asymmetric bell-shaped velocity profiles.