To address the issues of low accuracy in assigning weights to landslide hazard evaluation indicators and the inability of comprehensive evaluation methods to account for indicator randomness and uncertainty,this study fo-cuses on landslide hazards in the Wenchuan—Lixian section,an improved analytic hierarchy process(AHP)and co-efficient of variation method are used to calculate both subjective and objective weights for the landslide hazard eval-uation indicators,additionally,a combined weighting approach is utilized to compute the composite weights of the in-dicators.Finally,a cloud model is employed for a comprehensive evaluation of landslide points and compared with other evaluation methods.The results indicate that the combined weighting-cloud model method more closely ap-proximates actual conditions,with 85%of the most likely landslide occurrences aligning with the real situation,this confirms that the proposed method is reasonable and feasible,providing valuable scientific guidance for landslide prevention and control in similar regions.