An optimized Radial Basis Function (RBF) neural network algorithm was presented to accurately estimate the State Of Charge (SOC) of lithium batteries in electric vehicles. The parameters and structure of RBF neural network are optimized by Particle Swarm Optimization (PSO) algorithm, and the width and center of basis function in RBF neural network are determined. According to the charging and discharging mechanism of lithium batteries, voltage (U), current (I), internal resistance (R) and temperature (T) of the influence factors of SOC are taken as input vectors to conduct simulation experiments in Matlab. Experiments show that this method can achieve accurate, fast and convenient SOC estimation of lithium batteries. The error between prediction results and actual measurement results is less than 4%, which meets the technical index requirement of SOC prediction error of 5%.It has certain practical application significance for the estimation of lithium batteries SOC of electric vehicles.