Riparian erosion is one of the major causes of sediment and contaminant load to streams, degradation of riparian wildlife habitats, and land loss hazards. Because of the complex interactions between the mechanisms that govern soil detachment, assessing erosion vulnerability at the watershed scale is challenging. The main objective of this study was to develop a methodology to identify areas along the riparian zone that are susceptible to erosion. The framework was developed by integrating a physically-based watershed model MIKE-SHE, to simulate the hydrologic variables, and a habitat suitability model, MaxEnt, to quantify the probability of presence of erosion. The changes in elevation (i.e., erosion) were localized based on two LIDARs (2009 and 2012) and used as sample locations in MaxEnt. The environmental features used as constraints in the estimation were land cover, soil, slope, and the hydrologic variables. The modeling framework was evaluated in the Fort Cobb watershed, Oklahoma. Results showed that the most vulnerable areas for very high erosion (1.38 m to 4.27 m) were located at the upper riparian zones of the Cobb and Lake Sub-watersheds. Approximative ly 80% of the riparian zone has a 30% or less probability to experience "very high" erosion while the remaining 20% has a probability of up to 70%. Furthermore, the results revealed that soil type and slope were the most important predictor of riparian erosion across the watershed. Soils found to be the most prone to "very high" erosion belong to hydrologic group B and C with high sand or silt composition. By being able to identify the most vulnerable areas for stream and riparian sediment mobilization, conservation and managerial practices can be focused on areas needing the most attention and resources.