A mathematic model for filtering reducer transmission robots which have the characteristics of parameter uncertaintiesexternal disturbancestrong couplingnonlinearities and multi-variable is establishedand then an adaptive RBF networks backstepping method is proposed to control the robot.RBF is employed to approach the nonlinear item of the filtering reducer transmission robot on lineand an adaptive RBF network controller is developed based on backstepping method which effectively suppresses the effect of parameter uncertainties and external disturbance etc.It is combined with Lyapunov stabilization analytical method to demonstrate the close loop system convergence.Simulation results verify that the proposed controller not only satisfies high performance position trackingbut also has good accuracy and robustness.