With the improvement of the accuracy requirements of high-precision long endurance inertial navigation system, the gravity disturbance has become the main error source of inertial navigation system.Is the key factor to improve the navigation accuracy whether it can be effectively compensated. The traditional compensation method based on spherical harmonic model cannot reflect the detailed information of the earth’s gravity field, and thus has poor compensation result for medium and short wave gravity disturbance components. In the traditional state estimation method, the Markov state model has poor accuracy in describing the long wave gravity disturbance component, so the compensation result of the long wave component is poor. The above methods can not compensate the actual gravity disturbance that has wide band with high precision. To solve this problem, a high-precision gravity disturbance compensation method based on spherical harmonic model and multi-sensor information fusion is designed in this paper. On the one hand, the method compensates the long wave gravity disturbance component by using the high calculation accuracy of the low-order spherical harmonic model in the long wave band. On the other hand, using strapdown inertial navigation/laser Doppler velocimetry/barometric altimeter to form a fully autonomous integrated navigation system, the residual medium and short wave gravity distur-bance component is established as a high-precision Markov model, so as to realize the state estimation and compensation of medium and short wave component. The simulation results show that the proposed gravity distur-bance compensation method can effectively improve the estimation effect of gravity disturbance, realize high-precision gravity disturbance compensation and have high navigation accuracy.