To solve the problem of high computational complexity and large mutua.coupling effect of polarized array, a polarized sparse algorithm based on three-cross dipoleco-prime array is proposed for direction of arrival (DOA) estimation, A virtual linear arraywith more array elements is formed, by making full use of all information of the threechannels of three-cross dipole, so that the uncorrelated signals become coherent signals. Then theorward ano backwarc spatia smoothing method is used to deconerence the covariance matrix otvirtual array. The algorithm can expand the array aperture and reduce the computational complexity.The effectiveness of the polarized sparse algorithm is verified by several groups of simulationexperiments. The results show that the algorithm has better estimation accuracy.