Abstract: Research on path optimization problem in manual order picking. Under the condition of ABC storage strategy and double regional warehouse layout, path model was set up. Use genetic algorithm to solve the optimal path, make simulation experiments, and compare the optimal solution of S-shape, return-shape picking path models. The simulation results show that when small picking number, the genetic algorithm has obvious advantages, when large picking number, the three methods tend to be more close. And the applicability of the genetic algorithm was affected by the warehouse layout. The optimized picking path saves operation time, speeds up the flow of goods within the warehouse, improves the efficiency of the whole logistics activity, helps logistics activities completed quickly, and enhances customer satisfaction.