Conventional H∞filter is much more conservative, as the filter parameters are set in the initial while Kalman filter requires the statistical properties of noise accurately. For the above limitations, this paper proposed an adaptive H2/H∞ filter for multi-sensor integrated navigation systems. The method derives the gain weight coefficient using the theory of matrix inequalities and estimation variance matrix based on the least trace criterion. The method can improve the accuracy and robustness of integrated navigation systems by adjusting gain weight coefficient automatically. Finally, the adaptive H2/H∞ filter method is compared with Kalman filter and H∞ filter on the test platform, Results show that the adaptive H2/H∞ filter can provide better performance of integrated navigation systems by combining the advantages of the Kalman filter and H∞ filter.