Abstarct: This thesis presents a novel method for controlling flexible-joint robot manipulators. In this approach, by compensating flexibility effect the robot turns to rigid robot and then control methods for rigid robots are applied. Feedback linearization and direct adaptive fuzzy control, baxsed on rigid model are designed with torque control strategy and simulated on three-lixnk articulated flexible-joint robot. In the following, voltage control strategy is introduced and feedback linearization baxsed on this strategy is explained. Then an adaptive fuzzy system is designed for compensating flexibility effect and feedback linearization baxsed on rigid model is implemented on flexible-joint robot. In addition, performance of controller is evaluated by simulation. At the end, an adaptive fuzzy system is designed for improving performance of system against uncertainties. In all of these methods, by mathematical analysis, it is proved that the proposed control laws guaranty stability. Voltage control strategy, in comparison with torque control strategy has simpler structure therefore, implementation is easier.