For several days in March of 1975, reflected solar radiation measurements were obtained from smooth and rough surfaces of wet, drying and continually dry avondale loam at Phoenix, Arizona, with pyranometers located 50 cm above the ground surface and a multispectral scanner flown at a 300-m heigth. The simple summation of the different band radiances measured by the mulitspectral scanner proved equally as good as the pyranometer data for estimating surface soil water content if the multispectral scanner data were standardized with respect to the intensity of incoming solar radiation or the reflected radiance from a reference surface, such as the continually dry soil. Without this means of standardization, multispectral scanner data are moste useful in a spectral band ratioing context. Our results indicated taht, for the bands used, not significant information on soil water content could be obtained by band ratioing. Thusthe variability in soil water content should insignificantly affect soil-type discrimination based on identification of type-specific spectral signatures. Therefore remote sensing, conducted in the 0.4- to 1.0-μm wavelength region of the solar spectrum, would seem to be much more suited to identifying crop and soil types than to estimating of soil water content.