In this study, an optimal algorithm for a train speed profile using GA (genetic algorithm) to obtain optimum energy efficiency is presented, and its effectiveness is shown by simulation results. Based on a method that simplifies the train driving modes between stations (maximum-power driving, coasting, maximum braking) and by controlling the coasting point, the fitness function is set so that the train can travel the distance between two stations within the defined target travelling time. Then the coasting points that satisfy both constraint elements (distance between stations and target travelling time) are determined by applying GA. Train performance simulation blocks and GA blocks are designed using Simulink.