The primary objectives of this study were (1) to apply recently developed quantitative vegetation analysis procedures to the problem of describing the forest vegetation of the central portion of the Great Smoky Mountains National Park, (2) to group samples into forest types based on the importance of a large number of taxa and to compare the results with studies using types defined by relative density or basal area of canopy dominants, (3) to assess and further define the relationships of vegetation pattern with elevation and with topographic characteristics, (4) to examine possible relationships between soil characteristics and vegetation pattern, and (5) to examine the successful status of the forest types. Data were analyzed from 266 sample locations ranging from 759 to 1585 m elevation in the central portion of the Great Smoky Mountains National Park, in the vicinity of Mt. LeConte, Greenbrier Pinnacle, and Thomas Ridge. The 266 sample soils were tentatively classified as: Typic Dystrochrepts, 51%; Lithic Dystrochrepts, 14%; Typic Haplumberpts, 15%; Umbric Dystrochrepts, 10%; Lithic Histosols, 4%; Lithic Umbric Dystrochrepts 3%; Fragmental, 2%; and Lithic Umbrepts, 1%. Sample linear correlations among the soil, site and vegetation characteristics were computed. The highest number of significant vegetation-soil correlations occurred with clay content of the A and B horizons and with phw. Most of the general vegetation characteristics were significantly correlated with microtopographic position. Canopy sample plots were grouped into forest types based on taxa and importance values with the aid of the agglomerative minimum dispersion clustering procedure. The types were: Spruce-Yellow Birch, Yellow Birch-Hemlock, Hemlock-Buckeye, Basswood, Northern Red Oak, Red Maple-Sweet Birch, Red Maple-Northern Red Oak, Yellow Poplar, Chestnut Oak, Oak-Pine, Table-Mountain Pine-Pitch Pine and Table-Mountain Pine. The discreteness of the plot groups (types) was tested by canonical analysis. Vegetation, site and soil characteristics of the 17 forest types were described. Relative densities of tree taxa in the canopy, sapling and seedling strata were compared to judge the successional stability of the types. Types which had no evidence of past disturbance appeared to be relatively stable, although periodic reproduction apparently had occurred in some plots. Acer rubrum, Quercus prinus, Q. rubra and Oxydendrum arboreum were the most common tree taxa which had replaced American chestnut. A "topographic site gradient" was constructed based on combinations of potential solar irradiation classes (based on aspect and slope angle) and microtopographic slope positions. Each position along this gradient was assigned a number, termed the "site gradient index" (SGI), which increased toward mesic sites. The samples of each forest type were plotted on axes of elevation and SGI. A composite diagram was made portraying the pattern of most of the types on the SGI -elevation axes. Observable patterns were noted when Umbrepts, Umbric Dystrochrepts and Lithic Dystrochrepts were plotted on the site diagram. Other characteristics showing patterns on the diagram were: percent clay in the B horizon, total vascular taxa, percent shrub cover, tree sapling density, and canopy basal area. The combined vegetation-site summary contained in the composite site diagram suggests that the Mt. LeConte area departs significantly in detail (if not in basic outline) from the mosaic chart of Whittaker. This suggests that further local studies are needed in the Park to further verify or redefine its outline and / or details.