To deal with the problem that it is difficult to predict acceleration information in the process of intercepting highly maneuvering target,a novel 3D intelligence guidance law is proposed by using adaptive RBF neural network and backstepping sliding mode control technique.Firstly,a 3D guidance model for maneuvering target interception is established based on the principle of zeroing the rate of Line-of-Sight (LOS),and a finite-time guidance law is designed by using backstepping sliding mode control algorithm.Then,the target acceleration is regarded as the uncertainties of the control system,which is online estimated and compensated for by using RBF neural network.Meanwhile,an adaptive switching gain is designed to restrain the chattering of the control system.Finally,the stability of the closed-loop control system is proved based on Lyapunovs direct method,and the effectiveness and superiority of the proposed 3D guidance law are verified via comparative simulations.