Landslide susceptibility evaluation at regional scale is commonly performed based dominantly on the analysis of geological and geomorphological conditions of historical landslide cases. The main content of this type of evaluation covers identifying key casual factors, their critical groupings and relative importance. The present study demonstrates an application of the above concept to a 90 km long segment of Jinshajiang River valley in China. Correlations of landslide occurrence with potential causative factors are derived according to interpretation of field investigation. Lithology, orientation of bedding planes, slope angle, stream action, rainfall and earthquake intensity are selectively recognized as identifiable/measurable causative factors to establish a factor domain. The membership grades, for field values of quantitative factors, to the susceptibility classes are determined based on the construction of fuzzy sets, while those for descriptive factors are assigned from a fuzzy score table. The analytic hierarchy process (AHP) is adopted for assigning weights to each individual factor. Subsequently, the evaluation is implemented in a GIS program IDRISI, where four classes of landslide susceptibility are identified and delineated in the subject area. The approach described in the present paper showed consistence with the nature and availability of data for evaluating landslide susceptibility at regional scale. The methodology presented can be effectively employed by relevant authorities to identify risky areas for dislocating major infrastructural project, and develop management strategies for land use.