Technical Abstract: Improvement of grazinglands can contribute to increased farm profitability and ecosystem sustainability by reducing machinery requirements, time needed and fuel costs for harvesting forages, applying fertilizers, and distributing manure. Site type is effectively fixed (e.g. climate and landscape position), while plant community composition is more amenable to modification by selective addition or suppression of species. Careful choice of species can enhance important ecosystem functions such as primary production, drought resistance and invasion resistance. Our objective was to identify the site and vegetation types found in pastures, and to examine the fidelity between those types. Inventory data on vegetation, soils, and landscape attributes were collected from 82 pastures on 26 farms from 1998-2005 (188 samples including repeats). These data have been used to develop parallel classifications of site types and plant communities through nonhierarchical clustering and classification tree analysis. Classifications were evaluated on both their statistical properties and their utility for pasture management. Ordination methods were used to investigate overall patterns of distribution and as a basis for visualizing cluster membership. Overall community composition was related to geographic location, slope, aspect, elevation, and to soil properties including pH and organic matter. There was a many-to-many correspondence between site and vegetation classifications, probably resulting from management influences. The classification of landscape-based site types and the inventory of associated vegetation types will be combined with data on plant traits and ecosystem functions to refine methods to evaluate existing pastures and design new pastures to more effectively meet economic and environmental goals.