A fusion inertial navigation location algorithm for MT2503 and MEMS sensors is presented to improve accuracy and continuity. MT2503 chip is the locating terminal of location algorithm which blends accelerometer sensors, gyroscopes, magnetometers, and other sensors. Accelerometer sensors are used to resolve step size, stride and stride frequency. Gyroscopes and magnetometers are used to recognize the offset of a locating terminal. Finally, an offset of the location terminal is added on the original location to obtain the real time location of the locating terminal. Experiments confirmed that the problem of accumulative errors in navigation solution of the MEMS location algorithm is solved well through zero velocity update and Kalman filter. The method improves the accuracy of indoor inertial navigation positioning algorithm based on MEMS sensors.