A novel function-link fuzzy cerebelarmodel- articulation-controller (CMAC) is developed in this study. It is a generalization of a fuzzy neural network and of a conventional CMAC. Then, a control system comprising a function-link fuzzy CMAC and a fuzzy compensator is proposed for multi-input multi-output (MIMO) nonlinear chaotic systems. The function-link fuzzy CAMC is used to mimic an ideal controller and the fuzzy compensator is designed to suppress the approximation error between the function- link fuzzy CMAC and the ideal controller. The on-line learning algorithm of the controller’s parameters is derived to improve the control performance. Moreover, the design of the fuzzy compensator can guarantee the system’s stability. Finally, synchronization of the unified chaotic system has been examined to illustrate the effectiveness of the proposed design method.