In this paper, a novel high-resolution Direction of Arrival (DOA) estimation is proposed, which is based on the high order power of the inverse of spatial covariance matrix and Root-MUSIC. The MUSIC method has received attention as one of the "Super-resolution" DOA estimation methods because of its brilliant characteristics. However, MUSIC also has the problem of large computational capacity. In order to reduce the computational complexity, the high order power of the inverse of spatial covariance matrix is employed to obtain noise subspace, then Root-MUSIC is used to estimate the directions of the signals, and finally a coefficient of distance is defined to eliminate the false directions. This method does not estimate the number of signals, and has no need to perform eigenvalue decomposition and the peak search of spectrum. The simulation results show that the new method achieves the performance of the MUSIC algorithm while reducing the computational cost.