The Loess Plateau stands as one of China's most landslide-hazardous regions. While previous studies have investigated landslide distribution patterns and identified aggregation zones across this area, most have neglected to incorporate the magnitude of landslide risk into their analyses. This study established landslide risk grades by integrating both casualty and economic loss indicators. The spatial distribution patterns of landslides with different risk categories were investigated through spatial autocorrelation analysis, relationship analysis between landslide risks and key topographic/geological factors, and landslide kernel density estimation. The results demonstrate that: (1) Loess Plateau landslides exhibit distinct risk-dependent clustering patterns; (2) high-risk landslides show strong spatial correlations with large elevation gradients, fault systems, and seismic activity; (3) the southwestern plateau represents a high-risk concentration zone, while eastern areas demonstrate significantly lower risk. Building on these findings, this study proposes a multi-level landslide hazard intensity classification framework. Implementing this framework, we delineated eleven distinct landslide risk zones across the Loess Plateau, each defined by specific risk levels and spatial aggregation characteristics. This classification system simultaneously captures both hazard intensity and spatial distribution patterns. The study provides critical insights for regional-scale risk assessment and delivers a scientific basis for targeted disaster mitigation strategies and sustainable land-use planning in this vulnerable region. (c) 2025 COSPAR. Published by Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.