ASTER images were used for creating the hyperspectral mosaic of Crete. The resulting mosaics were processed intensively in order to detect lineaments related to faults. The particular data were compared to those coming either from surface observations or digitization techniques of geological maps. In addition to the above, a 20m SPOT derived DEM was employed along with other information layers (geology, climatic, landuse, etc) to classify areas with potential landslide risk. The model was based on weighted variables through statistical analysis of inter-related parameters.