Abstract: In order to reduce the influence of azimuth gyro drifts on the long time positioning accuracy of single axis rotational strapdown inertial navigation system (SINS), an online azimuth gyro drift estimation method is proposed. Error parameters of SINS are analyzed here, and it is pointed out that the precision of the eastward drift and azimuth misalignment determines the estimated precision of the azimuth gyro drift. Then, an optimized Kalman filtering algorithm is used to estimate the misalignment angles of the SINS and to compensate the misalignment angles errors of the system. Finally, the Kalman filtering algorithm is used again to estimate the inertial measurement unit (IMU) error on the basis of the first Kalman filtering step. The turntable three axes swinging and vehicle running experiment are carried out, and in the vehicle verification experiment, the inertial navigation positioning error is 2.12n mile within 12h after compensation. The positioning accuracy of the navigation system meets the long time navigation requirements of the medium precision singleaxis rotational SINS.