Fuels maps are a fundamental part of fire management activities such as prescribed fire planning, suppression strategies, smoke management, and fire effects. The constraints imposed by fiscal and human resources make it desirable to have a method that can rapidly and objectively map fuels over very large areas. The Multi-Resolution Land Characteristics Interagency Consortium personnel are producing a 30 meter resolution land cover data set for the conterminous U.S., based on a mosaics of Landsat Thematic Mapper (TM) images of both leaf-on and leaf-off periods. This spatially consistent mapping of the U.S. produces land cover classes that are too broad for any fire fuels mapping effort. The data set does, however, provide a mapping base upon which further land cover definition may be derived. A demonstration centered on the Big Horn Mountains of Wyoming, was produced with an unsupervised clustering of the original TM data to create subclass segmentation of representative original land cover classes. The mapping process demonstrated a simple, repeatable, and extendable methodology for extracting subclass spatial patterns of similar vegetation from a National Land Cover Data set. © University of Idaho 2000. Abstract reproduced by permission.