The invention discloses a neural network and active disturbance rejection composite control method for a three-axis tripod head servo motor. The method is characterized by comprising following steps of (1) establishing a three-axis tripod head internal frame servo system dynamical equation with a Stribeck friction model, (2) designing an active disturbance rejection controller for the dynamical equation established in step (1), and (3) designing an RBF neural network, combining the RBF neural network with the active disturbance rejection controller to form an RBF neural network active disturbance rejection supervisory controller. According to the method, better response speed and control accuracy can be achieved in a nonlinear system, the stability and robustness of a control system can beensured, the introduction of neural network can improve the adaptive ability of the system, and the method can adapt to a larger range of a controlled object under the condition of satisfying controlperformance requirements.