In this paper, a novel dynamical model is proposed for the multi-input multi-output electrically driven robot manipulators, by an observer-based robust adaptive fuzzy controller. The proposed control scheme utilizes current control effort, which is more efficient than the torque control approach. The proposed method is very simple, accurate and robust. Based on the adaptive fuzzy system an observer-based estimator is presented that uses feedback error function as the input of fuzzy system to approximate and adaptively compensate the unknown uncertainties and external disturbance of the system under control. Although the proposed controller scheme requires the uncertainties to be bounded, it does not require this bound to be known. An H_∞ robust controller is employed to an attenuate the residual error to the desired level and recompenses the both fuzzy approximation errors and observer errors. The proposed method guarantees the stability of the closed-loop system based on the Strictly Positive Real (SPR) condition and Lyapunov theory. The proposed control scheme is not limited only for controlling of robotics vehicles, it can be applied for a class of nonlinear MIMO systems. Finally, in simulation study, to demonstrate the usefulness and effectiveness of the proposed technique, a two-link robot manipulator system is employed.