In order to solve ESM-radar track correlation problem under condition of the influence on the maneuvering target close and cross, this paper presents a track correlation algorithm based on the improved Fuzzy comprehensive evaluation. In consideration of distance, angle, speed, phase difference and time difference information of the ESM sensor and radar in the cooperative passive location, the algorithm enriches the fuzzy factor set, gives out a dynamic calculation method of membership degree and weight in fuzzy comprehensive evaluation, and correspondingly improves the evaluation set and the evaluation criterion. The simulation results show that when the target is at the distance of 200m, the traditional multi-factor fuzzy integration judgment algorithm’s correlation accuracy is 55.2%, and this paper’s algorithm’s correlation accuracy is 74.0%. Under condition of increasing the target distance and the radar detection error, this algorithm’s correlation accuracy is always higher than that of the traditional multi-factor fuzzy integration judgment algorithm.