Abstract: The MEMS bionic vector hydrophone, which has advantages of high sensitivity, broad frequency band, vector and high Signal to Noise Ratio(SNR), is one kind of underwater acoustic signal detection device integrating piezoelectricity and MEMS technology. However, to make a further step in improving the predictive accuracy and the resonant frequency of underwater acoustic signal, optimization design of the MEMS hydrophone bionic microstructure is performed with finite element method in this paper. Firstly, it can be learned from theoretical formulas of natural frequency and stress that the natural frequency of bionic microstructure is inversely-proportional to the height of bionic cilia and the beam length, at the same time, is proportional to the beam width and thickness; Instead, the sensitivity is proportional to the bionic cilia height and beam length, and is inversely-proportional to the beam width and thickness. Based on the theoretical analysis, maximum stress curves and resonance frequency curves of sensor under different structural parameters are drawn. Secondly, a static analysis was done with ANSYS software and a response curve of natural frequency and stress was drawn. Finally, the simulation results show that a better performance of high sensitivity and broad frequency band for MEMS hydrophones can be obtained by designing beam length, width, thickness and bionic cilia height and radius as 400, 80, 50, 1000 and 80 m respectively. The simulation results and the theoretical analysis are compared, and the differences between them are analyzed.