Landslide is one of the destructive erosion events in slopes that causes financial losses on the roads. Identification of factors affecting landslide occurrence and hazard zonation mapping is considered as one of the basic tools to manage and reduce possible damage. In this research, we tried to zonate the landslide hazard at proposed Torghabeh- Darrod road by using Bayesian method, so that we can prevent many human and financial risks. Therefore, by collecting the required information on the landslides distribution in the area, the maps of landslide distribution, along with13 data layers including: slope gradient, aspect, curvature, elevation, landuse, geology, distance from road, distance from drainage, distance from fault, soil and rainfall zones was produced in the GIS environment. By using Bayesian probability theory, the relationships between each factors and landslide points were determined and subsequently the classes weights of each factors were obtained. Finally, landslide hazard zonation maps was prepared by Bayesian theory for the study area, in such a way that all effective factors were used and each factor was deleted. Finally the accuracy and precision of the maps were calculated using the ROC carve and 30% of landslide points. Data show that the precision of prepared probabilistic model by the Eleventh approach modeling (the distance from the fault was removed) was determined about 86.5% (very good) in study area. Using the provided map, we can identify unstable areas, and rely on them in the implementation of development planning particularly road construction.