The invention provides an optimal configuration method for feeder terminals based on an genetic algorithm, which is characterized in that an objective function is built according to the total investment of two-remote feeder terminals and three-remote feeder terminals under constraint conditions of a power supply reliability index of a power distribution network and power failure time of a load point, reliability evaluation is performed by adopting a fault consequence analysis method, influences imposed on the reliability of a power distribution system and the power failure time of a load pointby the two-remote feeder terminals and the three-remote feeder terminals are sufficiently considered, and finally the objective function is solved in an optimized manner by adopting the genetic algorithm. The optimal configuration method can solve a configuration problem in number and position of the two-remote feeder terminals and the three-remote feeder terminals, and provides a reference for scientifically and reasonably performing power distribution automation plan design and transformation.