Abstract: It is usually assumed in flowshop scheduling problem that all of the processing procedures are performed in one factory, which is termed as single factory production mode. Nowadays, the distributed manufacturing, due to its low cost, low risk, and high quality, has been becoming the main development trend in the industrial production. Therefore, it is necessary to study the distributed scheduling problem. The distributed assembly permutation flowshop scheduling problem plays an important role in modern supply chains and manufacturing systems. In this paper, a modified biogeography-based optimization (MBBO) algorithm is proposed for solving distributed assembly permutation flowshop scheduling problem to achieve the goal of minimizing makespan. In the initialization phase, the proposed MBBO uses SPT and NR₂ rules to optimize the feasible solutions. Then, in the mutation phase, the job insertion heuristic method based on the factory completion time is used to adjust the factory assigning and the processing order of the jobs. Moreover, the simulated annealing algorithm is combined to avoid the local optimal solutions and enhances the global search ability. Finally, simulation experiments are made via 900 small instances and 540 large instances, and the comparison is provided with the existing 12 heuristic and metaheuristic algorithms as well as the basic biogeography-based optimization algorithm. These results verify the superiority of the proposed MBBO. Meanwhile, the new best known solutions for 70 instances are found.