By considering the urban entironment effect and remote sensing interpretation ability synthetically, the urban entironment basal status was classified into eight classes. Taking the Shanghai within outer line as a case, information extraction methods of urban entironment basal status from remote sensing images were studied. By taking TM and SPOT fusion image as the remote sensing data sources, and based on the knowledge of eight classes identification from high resolution remote sensing images with eyes in the GIS database, classification and regression tree recognition methods were built for different types of urban entironment basal status from the fused images and the fusion data ratio images. Four classes information including residential areas, water bodies, urban virescence and agriculture were extracted from the remote sensing data for a large area of city.