A major impact of the invasive Myocastor coypus in the introduction range is the collapse of riverbanks and nearby infrastructure, such as railway lines, due to the species' burrowing activities. As the ubiquitous implementation of preventive measures along watercourses is unfeasible, identifying susceptible areas is key to guide targeted management actions. This study used species-habitat models to (i) identify the local environmental features of the railway line/watercourse intersections (RLWIs) that make them particularly susceptible to coypu damage, and (ii) predict species occurrence probability in a wide lowland-hilly area of northern Italy (Lombardy) to identify priority areas for monitoring. The local scale models stressed that the RLWIs most susceptible to burrowing were those surrounded by arable lands with interspersed hedgerows locally characterized by high herbaceous vegetation and clay soil. In urbanized and intensive agricultural areas coypu dens were generally located significantly closer to the railway, increasing the collapse risk. The landscape-scale species distribution model showed that lowland areas located along major rivers and lake shores, but also agricultural areas with a dense minor hydrographic network especially in the southeast of the study area, are more likely occupied by the species. The local scale models shown that specific environmental characteristics increase the risk of burrowing near RLWIs. The landscape scale model allowed predicting which areas require thorough monitoring of RLWIs in search of such local characteristics to implement preventive management measures. The proposed model-based framework can be applied to any geographical context to predict and prevent coypu damages. This article is protected by copyright. All rights reserved.