Maintenance service network is built by the manufacturer to provide timely maintenance and ensure reliable operation for machinery. It is of great strategic significance to design a reliable maintenance service network. Especially in the “three summer” and “harvesting, ploughing and sowing” period in agriculture, cross-region operation had become an important agricultural machinery service mode, which bring a great challenge to the maintenance of agricultural machinery. Thus, focused on dynamic moving and geographical distributed agricultural machinery, the joint optimization problem of the static and dynamic service facilities location and service region districting was proposed. A novel mixed-integer nonlinear programming model-based decision model was developed, and an effective solution algorithm integrating linearization method and Benders decomposition was designed to determine the location of static service stations and dynamic service vehicles. The service region districting problem was addressed simultaneously to assign service units to these selected service stations. Finally, a real case study in Henan Province was conducted to verify the performance of the developed mathematical model and proposed solution algorithm. Computational results showed that the optimal number of dynamic service vehicles was 26 and the total service costs were increased with the increase of number of dynamic service vehicles when the number was greater than 26. 