Medical robots, with manipulator arms having multiple degrees of freedom, play a crucial role in surgery for complex surgical tasks performed by instruments mounted to it. This thesis proposes and implements a tracking control of three-degree-of-freedom medical assistant robot manipulator. However, controlling the robot manipulator is difficult due to the nonlinear, coupled and complex natures of the robots manipulator dynamics. The research work focused on developing an intelligent control system, Artificial Neural Network (ANN) controller, that provides smooth motion tracking for it. The proposed ANN controller is implemented and tested using MATLAB/Simulink. The performance of controller is evaluated using the Root Mean Square Error (RMSE) and Integral Time Absolute Error (ITAE) metrics. We compare the performance of the ANN controller with manually tuned PID and genetic algorithm tuned PID (GAPID) controller. The robustness of the controller is tested by introducing external time-varying and constant torque disturbance. Simulation results show that the proposed controller effectively tracks reference trajectories with minimal error as compared with PID and GAPID. Moreover, the ANN controller result shows promising performance in the presence of external disturbances. Hence, replacing the GA-PID controller by ANN controller enhances the disturbance rejection ability of the GA-PID controller. Keywords: Artificial neural network, Genetic algorithm tuned PID (GA-PID) controller, Minimal invasive surgery, 3-DOF medical assistant robot, Trajectory tracking.