Abstract: According to the prematurity and easily trapping in local optimum by using particle swarm optimization (PSO) in the pattern synthesis of antenna arrays with sidelobe reduction and nulls control, a multiple subpopulation PSO algorithm (MSPSO)is presented in this paper. The MSPSO algorithm is built by employing the strategy of hierarchy and subpopulation with neighbor structure of von Neumann to improve the convergence speed and accuracy. A modified objective function model that adopts different fitness functions according to the character of the top and bottom layer is proposed to balance effectively the local and global searching ability of MSPSO algorithm. The simulation results show that the MSPSO algorithm can achieve relatively high performance in the pattern synthesis of antenna arrays.