Nowadays, air springs are gaining more popularity because they can significantly improve the ride comfort of vehicles. However, it is challenging to control an active suspension system with an air spring due to the strong nonlinearity. This paper proposes a radial basis function neural network (RBFNN) based adaptive sliding mode controller for the active suspension considering the nonlinear air spring. This method uses the universal approximation characteristics of RBFNN to estimate the nonlinear force of the air spring acting on the vehicle body. By designing a suitable Lyapunov function, the adaptive rate of the controller can be obtained, and the stability of the system can also be guaranteed. The proposed method can obtain better control performance and does not require an accurate air spring model, thus reducing the design difficulty of the nonlinear controller. The simulation results demonstrate that, compared with a traditional sliding mode controller (TSMC) and a passive suspension, the adaptive sliding mode controller has the lowest magnitude of sprung mass acceleration, indicating ride comfort improvement.