Abstract: The Renzong Subdistrict in Xi'an City is a typical mountainous township characterized by complex geology, frequent human engineering activities, and recurrent landslide and collapse disasters. To reasonably predict the spatial distribution characteristics of geohazards in mountainous townships, this paper selects Renzong as the study area for the optimization methods of geohazard susceptibility assessment in mountainous townships. Based on the large-scale geohazard survey and analysis of disaster-pregnant conditions, taking grid cells as the evaluation units, nine impact factors, including slope gradient, slope height, slope aspect, slope type, engineering rock group, slope structure, distance from roads, slope cutting degree, and vegetation coverage, are selected for geohazard susceptibility assessment in the study area with composite index method. Besides, the deformation zones and geohazard sites identified by field survey are rasterized and quantified to optimize the evaluation results of composite index method, with validation by ROC curves. The results show that the evaluation precision has increased from 81.7% to 92.4%, effectively enhancing the accuracy of geohazard susceptibility assessment for townships. The method provides a scientific basis for geohazard prevention and mitigation.