Mineralogical differences between altered rocks and most unaltered rocks in south‐central Nevada cause visible and near‐infrared (0.45 to 2.4 μm) spectral‐reflectance differences which can be used to discriminate them broad categories of rocks in multispectral images. The most important mineralogical differences are the increased abundance of goethite, hematite, and jarosite, and the presence of alunite, montmorillonite, and kaolinite in the altered rocks. Analysis of reflectance spectra recorded in the field showed that the altered rock spectra are characterized by broad absorption bands in the 0.45–0.50 μm and 0.85–0.95 μm regions which are due to electronic processes in the iron ions, and a band near 2.2 μm which is due to vibrational processes in the OH ions. These features are absent or weak in most of the unaltered rock spectra. Therefore, the shapes of the 0.45–2.4 μm spectra for these altered and unaltered rocks are conspicuously different. However, because of the wavelength positions and widths of the Landsat Multispectral Scanner (MSS) bands, these spectral differences are not apparent in individual or color‐infrared composite MSS images. The technique developed to enhance these subtle spectral differences combines ratioing of the MSS bands and contrast stretching. The stretched ratio values are used to produce black‐and‐white images which depict materials according to spectral reflectance; ratioing minimizes the influence of topography and overall albedo on the grouping of spectrally similar materials. Color compositing of two or more stretched ratio images to form color‐ratio composites provides additional enhancement. The most effective color‐ratio composite for discriminating between the altered and unaltered areas, as well as among many of the unaltered rocks in south‐central Nevada, was prepared using the following diazo color and stretched ratio image combinations: blue for MSS 4/5, yellow for MSS 5/6, and magenta for MSS 6/7. Altered areas appear green and brown in this combination. Field evaluation of this color‐ratio composite shows that excluding alluvial areas, approximately 80 percent of the green and brown color patterns are related to hydrothermal alteration. The remaining 20 percent consists mainly of pink hematitic crystallized tuff, a result of vapor‐phase crystallization, and of tan and red ferruginous shale and siltstone. Discrimination of this unaltered tuff from the altered rocks may be possible in the 2.2 μm region because this absorption band is absent in the tuff spectra. However, because the shale and siltstone are mineralogically and spectrally similar to the altered rocks, there appears to be little prospect of distinguishing these rocks from altered rocks in visible and near‐infrared multispectral images.