Abstract: A Kalman filter with kinematic motion constraints is presented to address the problem of increasing position error from INS on s vehicle when the GNSS/INS integrated navigation system is without a GNSS signal. By using the inherent motion constraints, such as approximation height constraint, velocity constraint, and attitude constraint, the freedom of motion of a vehicle will be controlled and the unknown model parameters will be reduced. Both motion constraints and new kind measurements will strengthen the solution in constraint Kalman filtering. The positioning accuracy of the GNSS/INS integrated navigation system can be improved when the GNSS signal is blocked, achieving continuous navigation.