Dynamic test data fitting method was widely used in Gyro dynamic north algorithm at present. In this study, rapid dynamic north method based on cross-correlation function was presented to overcome the shortcomings of the weak denosing effect and large amount of data in dynamic test data fitting method and to deduce find north principle formula. This method firstly modulated the output signal of a ring laser gyro and accelerometer to create a certain frequency triangular amplitude modulation wave by continuous constant speed rotation. Then, calculated the cross-correlation function of two lines with the same frequency reference signal and the output signal of gyro and accelerometer separately, based on the nature of the cross-correlation function associated with the frequency, which to eliminate the accuracy effect of find north by inertial device drift and noise. The simulation results show that the new algorithm can realize all-attitude north finding in high precision and the error of azimuth angle and attitude angle is less than 0.01° after 30 s find north. The experiment on the north-finder results indicates that this algorithm can effectively restrain all kinds of noise in the measurement process of inertial devices, and the standard deviation is about 32.7″ for 5 min, which can basically satisfy the requirement of high precision north-finding.