Regional air quality models, such as USEPA's Community Model for Air Quality (Models-3/CMAQ), must handle numerous processes that require gridded vegetation cover data. These simulated processes include surface heat and moisture fluxes, dry deposition, and biogenic emissions. Characterizing vegetation cover can be quite important for some of these processes. For example, the rate of emitted isoprene varies by several orders of magnitude depending on the distribution of tree species. In this paper, we describe a 1-km resolved vegetation cover data base that has melded the following components: (1) 1-km resolved broad classes from the USGS EROS land cover characteristics data base, (2) county-resolved crown cover for individual tree species from the US Forest Service's Forest Inventory and Analysis Data Set, (3) county-resolved individual crop cover from the 1992 US Agricultural Census, and (4) urbanized area definitions from the 1990 Census. The usefulness of this improved vegetation cover database is demonstrated with spatial estimates of biogenic volatile organic compound emissions (emitted mostly from forests) and biogenic nitric oxide emissions (emitted mostly from heavily-fertilized agricultural soils).