Diatom composition of four Lake Erie estuaries was related to seasonal factors, year, location within the estuaries, and water quality parameters including nutrient and metals concentrations. Canonical correspondence analysis (CCA) revealed seasonality as the most important factor determining variability in diatom species composition among sites and dates. Alkalinity, pH, silicate, orthophosphate, and nitrite concentrations were water chemistry parameters correlated with diatom community composition. Eigenvalues for the first two CCA axes of nutrient/physical data and species data were higher than the first two CCA axes of metals data and species data. In addition, the water quality of these estuaries was evaluated using an index composed of Lange-Bertalot pollution tolerance values. The Lange-Bertalot index scores indicated that the Ashtabula estuary had the best water quality of the study sites. Lange-Bertalot index scores were highly correlated with a gradient of disturbance represented by the first axis of a principle components analysis of sites and nutrient data (Spearman ρ = 0.7). The Lange-Bertalot tolerance values could be useful for discriminating ‘good' sites from “bad” sites among the Lake Erie estuaries.