Based on the refined 1∶10,000- scale geological hazard survey data for Zhenning County, Gansu Province, including hazard points, slope units, and remote sensing interpretation points, the study systematically analyzes the predisposing conditions, formation mechanisms, and disaster patterns of typical hazards, aiming to elucidate the deformation characteristics and evolutionary trends of landslides, collapses, and other geohazards within the study area, thereby providing a scientific basis for local ecological migration and hazard- avoidance relocation, disaster prevention and mitigation, and the subsequent implementation of dual- control measures for hazard and risk.As a result, the primary landslide types identified in the study area include excavation- induced landslides, rainfall- triggered landslides, river lateral erosion- induced landslides, and loess—bedrock composite landslides, exhibiting characteristic failure patterns of “retrogressive tensile- cracking”, “creep- pushing”, “creep deformation”, and “soil—rock composite”, respectively. The predominant collapse types are “toppling rockfalls” and “sliding rockfalls”, with their associated failure patterns characterized as “falling + impact destruction + burial” and “impact- scouring destruction + burial”, respectively.Ground subsidence in the study area primarily manifests as cave dwelling collapse, while valleys exhibit a water- triggered loess geological hazard chain characterized by rockfall/landslide debris flow/mudflow. The disaster formation process progresses through the following sequence: human improper activities rainfall infiltration loess collapse, cracking, and sinkhole development loess slope subsidence/collapse or landsliding valley mudflow destruction and burial of hazard bearing body.The findings elucidate the deformation and failure mechanisms governing landslides, collapse, cave dwelling collapses, and valley hazard chains, thereby facilitating future rapid identification of potential hazards, precision- based implementation of monitoring and early warning systems, and rational planning of evacuation routes.