Abstract: To solve the transfer alignment problem on moving base of inertial navigation systems (INS), velocity plus attitude plus rate matching method was presented. Making use of three groups of parameters from both master and slave INS as the observations, Kalman filter was applied to estimate states such as the misalignment angle and the fixed installation error rapidly and exactly for the precise initialization of the slave INS. A statistic structural flexure movement model was introduced. According to the basic principle of transfer alignment, the state equation and the measurement equation were given in detail. Simulation results under the same conditions show that this scheme is as accurate as the velocity plus attitude matching and velocity plus rate matching. Besides, its evaluation time is only half of the other two's. It can be used on airborne and seaborne tactical missiles to reduce the navigation bias and guidance error effectively.