Coal-mining impacts on Smoky Creek, eastern Tennessee were evaluated using water quality and benthic invertebrate data. Data from mined sites were also compared with water quality and invertebrate fauna found at Crabapple Branch, an undisturbed stream in a nearby basin. Although differences in water quality constituent concentrations and physical habitat conditions at sampling sites were apparent, commonly used measures of benthic invertebrate sample data such as number of taxa, sample diversity, number of organisms, and biomass were inadequate for determining differences in stream environments. Clustering algorithms were more useful in determining differences in benthic invertebrate community structure and composition. Normal (collections) and inverse (species) analyses based on presence-absence data of species of Ephemeroptera, Plecoptera, and Tricoptera were compared using constancy, fidelity, and relative abundance of species found at stations with similar fauna. These analyses identified differences in benthic community composition due to seasonal variations in invertebrate life histories. When data from a single season were examined, sites on tributary streams generally clustered separately from sites on Smoky Creek. These analyses compared with differences in water quality, stream size, and substrate characteristics between tributary sites and the more degraded main stem sites, indicated that numerical classification of invertebrate data can provide discharge-independent information useful in rapid evaluations of in-stream environmental conditions. (Author 's abstract)