Abstract: For the estimation of low-altitude high-speed flying target trajectory, a method of multi-array space matching is proposed. In this method, the space points within the detection range are first divided, and then the true azimuth and elevation angles relative to the array reference coordinates are calculated to build a dictionary for storage. The signal received by the sensor array is processed in frames, and the direction of arrival is calculated for each frame of data. Theoretically, within the detection range, the high-speed flying target moves in a straight line. In the ideal condition, the direction vector of a single array obtained from each frame of data and the straight line formed by the reference coordinates are in the same plane. By matching the direction of arrival estimated by each frame of data in the dictionary, the N space points with the smallest error are found to fit a straight line. Then, the spatial plane is fitted by the straight lines obtained from more than two frames of data to obtain the estimated plane. The intersection line of the estimated planes obtained by the two arrays is namely the estimated high-speed flying target trajectory. Simulation results verify the effectiveness of the algorithm for the trajectory estimation of high-speed flying targets.