Abstarct: Robotic manipulators driven by electric motors without using gears are called direct drive robots. These robots show a better performance due to eliminating backlash, compliance and friction caused by gear transmissions. Dynamics of manipulator is very complicated, extensively computational and uncertain. On the other hand, fuzzy control can play an important role because of having knowledge baxsed rules. It can be easily designed for systems with an unknown structure. In this thesis, a fuzzy Proportional - Derivative control is designed for a Puma robot arm. The robot is driven by permanent magnet dc electric motors for tracking position of a desired trajectory. Studies show that fuzzy controller is successful in controlling the nonlinear behavior and overcoming uncertainty. However, to ensure stability and provide good performance require the use of all state variables. Computational burden due to the large number of inputs and fuzzy laws will be greatly increased, thus makes it difficult to implement the control. This thesis presents a new control scheme to reduce the computational burden and improves the performance of fuzzy control system. The proposed controller consists of a fuzzy proportional-derivative control baxsed on voltage control strategy with compensated current and feed-forward of the desired joint speed. Stability of control system is guaranteed. Mathematical analysis and simulation results show that the effectiveness of the proposed approach. In addition, its performance is better than the performance of fuzzy proportional- derivative control.