In actual production, production scheduling and equipment maintenance affect each other, and they should be optimized together. In order to study the distributed flexible job-shop scheduling problem with preventive maintenance, a hybrid genetic algorithm with two-population optimization mechanism is proposed to minimize the maximum completion time. Combined with the characteristics of the problem, the 3-D encoding and the corresponding decoding scheme are designed. Different strategies are used to initialize the population to balance part of the factory load. Different crossover and mutation operators are designed for the two populations to improve the diversity of the algorithm. The cooperative optimization of two populations is realized using the method of exchanging elite solutions. Local searches are designed for critical factories and preventive maintenance operations. Finally, compared with the existing algorithms, experiments on homogeneous and heterogeneous factories are carried out, and orthogonal test method is used to optimize the parameter setting of the algorithm. The simulation results verify the effectiveness of local search and population cooperation and the superiority of the two-population hybrid genetic algorithm to solve the distributed flexible job shop scheduling problem with preventive maintenance.