According to the traditional adaptive polarization filter algorithms had problems of low convergence speed and the iterative factor being hard to choose, a fast adaptive polarization filter was designed by utilizing the estimation theory of polarizing cluster center. The interference signal in polarimetric radar reception could be cancelled accurately and adaptively pulse-by-pulse. The range-gated technique in-pulse was adopted to obtain the interference signal samples. The current polarization state of the interference signal was estimated rapidly by computing the polarizing cluster center directly. Based on the principal of minimum interference output power, the fast filter process could be carried out. Compared to the traditional polarization filter, it is faster in convergence and more stable to the performance of interference suppression. Rapidity and efficiency of the filter algorithm are validated by simulation results.