The area under study is located within the Mekele out lier. It comprises Mesozoic sedimentary rocks unconformably overlain by trap volcanics. Landsat-TM images have been processed and geologically analysed in combination with revision of the previous works and present field survey for lithological mapping and reconstruction of the stratigraphy of the study area. In remote sensing the sensor and the obj ect to be sensed are located remotely apart. The link is due to electromagnatic (EM) waves. Thus in image processing and interpretation the knowledge of how EM radiation is generated, propagated, and modified when it interacts with materials is the main key. Materials produce a diagnostic spectral signature in response to the EM interaction. Generally, spectral signatures are related to the nature of the corresponding rock, which help to identi fy a material through image processing. Image processing includes statistical and graphical evaluation, ratioing, contrast enhancement, principal and canonical component analyses, combination of selected bands, texture and vegetation index analyses, classification, and visual interpretation of the images. Structures and other lineaments, lithological boundaries, and drainage patterns are traced directly from the screen during image interpretation. Ten mappable units are recognized both by digital image processing and visual interpretation. These are from youngest to oldest: Dolorite, Basalt, Amba Aradam Formation, Agula Shale, five separately mappable Antalo Limestone units, and Adigrat Sandstone. From sedimentary point of view, Adigrat Sandstone is fluviatile in origin, while the Antalo Limestone and Agula Shale are related with transgression and regression of the sea during Jurassic time. Amba Aradam Formation, on the contrary, lying unconformably on Agula Shale, is probably due to the withdral of the sea caused by an uplift of cratonic area related to the separation of America and Africa plates during the Cretaceous time. The final thematic images are geometrically rectified using selected ground control points to refer them with other maps. Digitally processed images expose fairly well the main lithologic units with some limitation.