The regional distribution of arsenic and 20 other elements in stream-sediment samples in northern Nevada and southeastern Oregon was studied in order to gain new insights about the geologic framework and patterns of hydrothermal mineralization in the area. Data were used from 10,261 samples that were originally collected during the National Uranium Resource Evaluation (NURE) Hydrogeochemical and Stream Sediment Reconnaissance (HSSR) program in the 1970s. The data are available as U.S. Geological Survey Open-File Report 02-0227. The data were analyzed using traditional dot maps and interpolation between data points to construct high-resolution raster images, which were correlated with geographic and geologic information using a geographic information system (GIS). Wavelength filters were also used to deconvolute the geochemical images into various textural components, in order to study features with dimensions of a few kilometers to dimensions of hundreds of kilometers. The distribution of arsenic, antimony, gold, and silver is different from distributions of the other elements in that they show a distinctive high background in the southeast part of the area, generally in areas underlain by the pre-Mesozoic craton. Arsenic is an extremely mobile element and can be used to delineate structures that served as conduits for the circulation of metal-bearing fluids. It was used to delineate large crustal structures and is particularly good for delineation of the Battle Mountain-Eureka mineral trend and the Steens lineament, which corresponds to a post-Miocene fault zone. Arsenic distribution patterns also delineated the Black Rock structural boundary, northwest of which the basement apparently consists entirely of Miocene and younger crust. Arsenic is also useful to locate district-sized hydrothermal systems an d clusters of systems. Most important types of hydrothermal mineral deposit in the northern Great Basin appear to be strongly associated with arsenic; this is less so for low-sulfidation epithermal deposits. In addition to individual elements, the distribution of factor scores that resulted from principal component studies of the data was used. The strongest factor is characterized by Fe, Ti, V, Cu, Ni, and Zn and is used to map the distribution of distinctive basalts that are high in Cu, Ni, and Zn and that appear to be related to the Steens Basalt. The other important factor is related to hydrothermal precious metal mineralization and is characterized by Sb, Ag, As, Pb, Au, and Zn. The map of the distribution of this factor is similar in appearance to the one for arsenic, and we used wavelength filters to remove regional variations in the background for this factor score. The resulting residual map shows a very strong association with the most significant precious metal deposits and districts in the region. This residual map also shows a number of areas that are not associated with known mineral deposits, illustrating the utility of the method as a regional exploration tool. A number of these prospective areas are distant from known significant mineral deposits. The deconvolution of the spatial wavelength structure of geochemical maps, combined with the use of large regional geochemical data sets and GIS, permits new possibilities for the use of stream-sediment geochemistry in the study of large-scale crustal features as well as the isolation of mineral-district scale anomalies. ?? 2006 Society of Economic Geologists, Inc.