In order to improve the dynamic pointing accuracy of electro-directional infrared targets during mount flight, under the condition that the the update frequency of position information is difficult to increase due to the hardware limit, a solution for the Kalman filter algorithm was proposed to predict the position information of the target and to increase the update rate. The model of dynamic pointing error based on factors such as flight speed, attitude, and sampling frequency of position information is established. Theoretical analysis and experimental verification were carried out. The results show that when the position sampling rate of the infrared target system is increased from 5 Hz to 10 Hz during the uniform linear flight, the pointing angle error is reduced by 54.66%. This study provides a useful attempt to reduce the dynamic pointing error of directional radiative infrared targets based on existing conditions.