To mitigate crop damage by insect pests, it is important to determine priority areas for the allocation of available pest management resources, which are usually limited. We tested whether the occurrence of pecky rice damage caused by the sorghum plant bug Stenotus rubrovittatus (Hemiptera: Miridae), a major rice pest in Japan, could be predicted using a spatial model based on land use data. Using a data from a 3-year field study, we examined the relationships among the land use of the area within a 300-m radius around each focal paddy field, the abundance of S. rubrovittatus, and level of pecky rice damage in the Maesawa region of northern Honshu Island, Japan. We also used mapping to visualize potential priority areas using a model and GIS software. From a linear mixed model analysis and model selection by Akaike's information criterion values, areas of source habitats, soybean fields and rice paddies were selected for the best model, but the abundance of S. rubrovittatus was not. Based on the model's evaluation, the predicted value of pecky rice damage, when compared with the observed value, was not sensitive enough for a quantitative prediction. However, the model was accurate enough to predict whether the brown rice was first grade, which is of greatest importance to local fanners. Therefore, it is possible that potential pecky rice damage by S. rubrovittatus could be predicted when the spatial arrangement of arable fields in a certain year is determined. Our results will be useful to support decision-making that involves insecticide applications to mitigate perky rice damage.