For the highprecision positioning and alignment of electrooptical aided inertial navigation system, the calibration of the sensitive axis installation angle between systems is the key to the initial alignment of the inertial navigation system(INS). Through analyzing the main influencing factors of the sensitive axis installation angle, the twoposition information is used to calibrate the sensitive axis installation angle without the aid of external positioning information. The algorithm of recursive leastsquares (RLS) parameter identification is used to accurately estimate the errors of inertial navigation system and the sensitive axis installation angle by analyzing the error of the model. The simulation results show that the estimation error of attitude angle is less than 0.1′,the calibration accuracy of the sensitive axis installation angle is better than 1′,and the algorithm is simple, easy to operate, less time consuming and independent of the external environment. Thus,the highprecision dynamic alignment requirement for the system can be satisfied.