It is worthy of further exploration to use spatial modeling in quantitative simulation analysis of the land price space. This paper took residential land in Nanjing as the research object, selected feature driving factors affecting the spatial distribution of residential land, based on spatial sampling and multiple regression models, established space factor and residential land price multivariate regression model, used raster GIS to simulate space of urban residential land price distribution. The results proved good simulation results achieved, which could reflect the real size of the impact of various types of features elements on the residential land price space, especially the rail traffic impact. Therefore, the simulation study of residential land price space could provide the reference value efficiently to improve the value of urban land and achieve urban land intensive and efficient use.