In the SINS / GPS integrated navigation system, the traditional unscented Kalman filter (UKF) uses the symmetric sampling unscented transform (UT), which can not meet the realtime requirement, and the distance from the sampling point to the center point will be increased as the number of state dimension increases, resulting in sampling of non-local effects. To solve the above problems, the minimum skewness simplex UKF reduced sampling strategy is used to improve the realtime performance of the system. Use the proportion of UT transformation to solve the nonlocal effects in the process of sampling, by adaptively adjusting the scaling factor to improve the estimation precision of the UKF. This paper introduces an improved UKF algorithm, adaptive scaled unscented Kalman filter (ASUKF) for SINS / GPS integrated navigation system. The simulation results show that this method has low computational complexity and high accuracy.