Nowadays, multi-robot systems are being utilized to perform agricultural tasks. It is particularly essential to robotize pesticide spraying because of the risk of poisoning workers. However, the problem is that the human-driven sprayers are big size, and difficult to maneuver in many orchards. Therefore, heavy spraying robots must be replaced with lighter robots. Also, it takes a lot of times to perform spraying using only one sprayer in a large orchard. To achieve this goal, the multi-robot system (MRS) can be applied to the spraying by robot cooperation to improve performance. In this study, we developed a task allocation system based on a Voronoi diagram for a multi-robot spraying system in an orchard. The seed point for area partition using the Voronoi diagram was obtained through node clustering using a k-means clustering algorithm. In the experiment, workspaces were partitioned according to the number of robots, from 2 to 10. A total of four metrics were used to evaluate the performance of the system. The results confirmed that our task allocation system is applicable to real orchards. INDEX TERMS Agricultural robot, multi-robot system, task allocation, area partition, unmanned ground vehicle (UGV).