Abstract: A variable structure neural network adaptive robust control (VSNNARC) is proposed for a class of uncertain nonlinear SISO systems. A neural network which is adjusted by using nodes activation and passivation techniques to minish the size of the neural network and computation load is used as an opproximator for the system unknown nonlinear functions. And the adaptive robust control is used for the weight learning and compensation to the modeling error and extern disturbances. The adaptation law of neural networks weights and the design method of robust controller are given out based on the Lyapunov stability analysis. Furthermore, the proposed controller can guarantee not only global stability but also transient performance. Finally, the controller is applied to a position tracking system of a turntable. The experimental results show that the system can perform good tracking and has strong robustness.