As projected climate changes loom, the monitoring of the response of natural vegetation becomes important for both science and management. Successful monitoring requires good baseline information and vegetation change detection techniques. The research reported here involved three main tasks: 1) digital geometric rectification of a detailed historic vegetation map; 2) an analysis of high spatial resolution airborne remote sensing data for tree mortality; and 3) the development of a Landsat Thematic Mapper-based vegetation change detection procedure. These studies focused on the Sierra Nevada of California, and in particular Yosemite National Park. The Vegetation Type Maps (VTMs) (Wieslander 1935) represent some of the finest maps of their kind in the world, and cover more than 40% of California. Yosemite National Park was mapped using these techniques in the late 1930s. Geometric inaccuracies in the 19th century USGS basemaps were mitigated using newly available GIS and remote sensing tools, enabling the rectified VTMs, to be integrated into the National Park's vegetation monitoring work. In 1992, several transects of a four-band high spatial resolution airborne scanner (ADAR) were taken of mid-elevation forests in the southern Sierra Nevada, to evaluate their use for tree mortality monitoring. This analysis highlighted the difficulty of using single-date imagery for monitoring vegetation changes, but showed 1) the best measure of tree mortality (when compared with field data) was found using solely the red wavelength band; and 2) the metric most highly correlated with the field data was relative canopy mortality (%), not absolute area (ha). Lastly, based upon Principal Components Analysis, I developed an algorithm for separating spectral changes resulting from vegetation changes on the ground from other changes present but not of concern to monitoring. Three vectors were derived using 6 of 7 Thematic Mapper bands. Although few published change detection methods have used the thermal IR band (6), it is an excellent means to detect changes in local evapotranspiration and thus vegetation. The results of this algorithm were quite good, with both the direction of vegetation change and its magnitude output. In general, areas of vegetation decrease were more easily detected than areas of vegetation increase.