Water erosion results in sedimentation, which reduces storage capacities in reservoirs and lowers their productive life. To effectively develop strategies for the mitigation of reservoir sedimentation, an assessment of the spatial variation of erosion is necessary. In data-scarce areas, soil erosion proxies such as aggregate stability can be used to map erosion hotspots. To assess the potential of using aggregate stability as a proxy for soil erosion, 90 sampling sites in the Ruiru catchment (Kenya) were selected using conditioned latin hypercube-based sampling. Aggregate stabilities were determined based on mean weight diameter (MWD). Thereafter, digital soil mapping (DSM) of MWD was applied to identify erosion-prone areas. Correlation analysis between MWD, soil properties and covariates revealed that organic carbon had the highest influence (27.9%) on MWD. When comparing the MWDs under different land-uses, areas under cropland and tea plantations had lower MWDs (2.54 ± 0.39 mm and 2.83 ± 0.36 mm, respectively) than forested areas (3.18 ± 0.09 mm) and were more susceptible to aggregate breakdown. A spatial map created using DSM revealed that earthen roadsides had the lowest MWD of 2.07 ± 0.27 mm; thus, highlighting their potential role in contributing to erosion. From our findings, the prediction of aggregate stability appears to be a valuable resource tool to identify erosion ‘hotspots’ and might be used for the development of catchment management plans aiming at mitigation of soil degradation.