This article demonstrates how multi-spectral Thematic Mapper data can be processed to enable discrimination of different surficial materials in the piedmont zone of the Tunisian Southern Atlas. The spectral reflectance characteristics of four types of piedmont surficial material; oppositional washes, stone pavements, sandy loams and gypsum crusts, are measured in the laboratory to determine the factors controlling their response in the Thematic Mapper bands. Three image processing techniques; contrast stretched false colour composites of band triplets selected by transformed divergence, ratio composites and decorrela-tion stretched false colour composites, are used to maximize the contrast between these surficial materials. These enhanced images can then be photo interpreted to produce geomorphological maps of the piedmont zone.