Abstract: The state of charge（SOC） of electric vehicle（EV）,which is one of the most important parameters of EV’s comprehensive performance, directly affects the vehicle’s driving range.Because there are many factors that affect the state of charge, and the vehicle driving conditions are complex and changeable, it is difficult to build an accurate mathematical model to analyze the state of charge.In this paper, the SOC prediction model of electric vehicle is built by using neural network, the battery voltage and vehicle speed are taken as the important parameters input model to predict the change of SOC parameters at the next time, through a lot of experimental data analysis and fitting calculation, it is shown that the proposed model has high precision and can be used to predict the SOC of EV.