Abstract: Design of the metamaterial microstrip antennas usually relies on experience, and the designs of metamaterial cells are mostly about size or shape optimization. In this paper, the influence on the gain performance of a microstrip antenna from the conventional metamaterial is studied, and it is found that its effect on the gain performance is limited. Then a topology optimization design method of the metamaterial microstrip antenna for high gain is presented based on the genetic algorithm. The design of metamaterial cells adopts the global design method; and taking the antenna gain maximization as the design goal, the presence or absence of square copper patch lattices as the design variables, the topology optimization model of metamaterial microstrip antenna within K-band (24 GHz) is established. Furthermore, based on the genetic algorithm, a novel metamaterial microstrip antenna microstructure configurations is obtained, simulation results show that the lateral radiation of the optimized microstrip antenna is suppressed, and its maximum gain is increased to 10.5 dB, the improvement rate is 35% compared with the case of an ordinary antenna. At the same time, the convergence of the optimization design problem is analyzed through changing the layout sizes of square copper patch lattices. The results show that the creative configuration design of the metamaterial microstrip antenna is convergent, and the microstructure configuration for 10*10 lattice size is the most cost-effective. Finally, the matching comparison on working frequency of microstrip antenna is studied between the metamaterial cells separate design and the global design,the results confirm that the global design of the metamaterial cells is necessary for the topology optimization design of the metamaterial microstrip antenna.