Abstract: The macro block-characteristic basis function method (MB-CBFM) is a numerical method for analyzing the electromagnetic characteristics of large-scale finite periodic array. In this technique, the CPU time and memory requirement are O(N), where N is the total number of unknowns of the array. The unit cells in the finite periodic array are divided into P categories of the macro blocks. Finally, a reduced matrix equation with PK unknowns can be obtained, where K is the number of the elementary basis functions on the block. Although the calculation of the reduced matrix uses periodicity, the impedance element is still calculated by the method of moments and the filling of the impedance matrix is time-consuming. In this paper, the fast dipole method is used to accelerate the generation of the reduced matrix. The fast dipole method expands the distance between the interacting dipoles with Taylor′s series, and the calculation efficiency of impedance elements is higher than that of the conventional method of moments. Numerical results show that the results of the proposed algorithm are in good agreement with those of the MB-CBFM and the FEKO. At the same time, the CPU time is significantly reduced.