An index of biological integrity (IBI) was developed for streams in the Hudson, Delaware, and Susquehanna River drainages in the northeastern United States based on fish assemblage data from the Mohawk River drainage of New York. The original IBI, developed for streams in the U.S. Midwest, was modified to reflect the assemblage composition and structure present in Mid-Atlantic Slope drainages. We replaced several of the Midwestern IBI metrics and criteria scores because fishes common to the Midwest are absent from or poorly represented in the Northeast and because stream fish assemblages in the Northeast are less rich than those in the Midwest. For all replacement metrics we followed the ecology-based rationale used in the development of each of the metrics of the Midwestern IBI so that the basic theoretical underpinnings of the IBI remained unchanged. The validity of this modified IBI is demonstrated by examining the quality of streams in the Hudson, Delaware, and lower Susquehanna River basins. The relationships between the IBI and other indicators of environmental quality are examined using data on assemblages of fish and benthic macroinvertebrates and on chemical and physical stream characteristics obtained during 1993-2000 by the U.S. Geological Survey's National Water Quality Assessment Program in these three river basins. A principal components analysis (PCA) of chemical and physical variables from 27 sites resulted in an environmental quality gradient as the primary PCA axis (eigenvalue, 0.41 ). Principal components analysis site scores were significantly correlated with such benthic macroinvertebrate metrics as the percentage of Ephemeroptera, Plecoptera, and Trichoptera taxa (Spearman R = -0.66, P < 0.001). Index of biological integrity scores for sites in these three river basins were significantly correlated with this environmental quality gradient (Spearman R = -0.78, P = 0.0001). The northern Mid-Atlantic Slope IBI appears to be sensitive to environmental degradation in all three of the river basins addressed in this study. Adjustment of metric scoring criteria may be warranted, depending on composition of fish species in streams in the study area and on the relative effort used in the collection of fish assemblage data.