Abstract: Target discrimination is a key technique in missile defense system, and after analyzing the defense system configuration, an integrated target discrimination mathematical model based on recursive temporal-spatial data fusion is performed, which brings the expert knowledge, environmental information and measure information into the integrated discrimination flow. During spatial fusion process, the reliability of discrimination approach is measured by Analytical Hierarchy Process (AHP), and the concept of fusion weight is introduced to measure the value of discrimination approach for participating in fusion, which is determined by reliability of discrimination approach and consensus among sensors results. During temporal fusion process, the fusion discrimination result is inherited and updated by Dempster-Shafer theory. Simulation results show that the integrated discrimination model can fuse the local decision from discrimination approaches effectively, restrain the influence of singularity and tolerate the erro=、r of discrimination.