Detrended correspondence analysis (DCA) was evaluated for its usefulness in elucidating relationships among samples and among species of periphyton in an oligotrophic stream, and for its effectiveness in displaying major gradients where an experimental gradient (copper) affecting species composition was imposed. It was highly sensitive to differences among samples and consistently provided ecologically meaningful species ordinations. Gradients related to seasonality of taxa and year-to-year differences in population densities were evident in DCA ordinations if data for all sampling dates were included, and these gradients complicated interpretation of the copper gradient. Stage of succession was a secondary gradient during exposure and complicated interpretation of the copper gradient after a major disturbance event (flooding).