Abstract: To optimize the layout of rural residential areas and promote the coordinated development of urban and rural areas, the Analytic Hierarchy Process (AHP) was used to evaluate the suitability of rural residential areas. Selecting Longxian in northwestern region as the research area, which had various landform types and complicated terrain, and based on the features of the county and the characteristics of rural residential land use, the evaluation indexes were determined from two aspects of natural factors and social economic factors. The elevation, slope, slope direction and water source influence were evaluated as natural factors, and the accessibility of roads, the accessibility of urban centers and the scale of land use were as socio-economic factors. With ArcGIS 10.3 software support, using the method of AHP, and taking the grid of 20 m×20 m as the evaluation unit, multi- factor comprehensive evaluation was used to estimate rural residential areas in Longxian. The results of suitability evaluation of residential land was graded into four classes of optimal, suitable, basically suitable and unsuitable, accounting for 13.88%, 42.84%, 34.48% and 8.80%, respectively. Optimal rural residential area had superior natural condition and convenient transportation, was close to the town center and large scale of land, and was suitable for the construction of facilities. Suitable rural residential area had wide distribution of land, slight fluctuation of terrain, convenient transportation, high accessibility of urban centers and a certain scale. Basically suitable rural residential area had slightly poor natural condition, and general accessibility of roads, and limited accessibility of urban centers and the scale of land use. Unsuitable rural residential area had poor topographic condition and inconvenient transportation, far from the center of the town in scattered plot, and it is not conducive to the development of residential areas. The results provide a theoretical basis for the selection of rural residential land.