To solve the problem of fault location of power distribution network, a fault location method based on fruit fly optimization algorithm is put forward. The continuous location coordinates and search steps of the original fruit fly optimization algorithm are discretized to link with the status information of feeder sections. Objective functions are established for radial distribution network with single power supply and complex distribution network with multiple power supplies. Four fault scenarios including single fault, single fault with information distortion, multiple faults and multiple faults with information distortion are tested respectively, and the test results validate the accuracy and effectiveness of the proposed method. Compared with bat algorithm, particle swarm optimization and genetic algorithm, fruit fly optimization algorithm turns out to be good at finding an optimal solution to multiple faults location problems with a faster convergence speed. This work is supported by National Natural Science Foundation of China (No. 51467012).