This data release contains predictions of selected  fish community metrics and fish species occurrence using Random Forest models with landscape data for inland reaches across the Chesapeake Bay Watershed (CBW). Predictions were made at four time intervals (2001, 2006, 2011, and 2016) according to changes in landcover using the National Land Cover Database (NLCD). The fish sampling data used to compute these metrics were compiled from various fish sampling programs conducted by state and federal agencies, county governments, universities, and river basin commissions across the watershed. Community metrics describe composition, tolerances, habitat preferences, and functional traits of fish communities (and were derived from Krause and Maloney, 2021). Community analyses were developed for four aggregated Level III ecoregions: Northern Appalachians (NAP), Southern Appalachians [split into two sub-regions; the SAP-Piedmont (Piedmont and Northern Piedmont Ecoregions; SAPPIED] and other ecoregions [SAPNW]) and the Coastal Plains (CPL), and a final index was calculated for each ecoregion as the average of selected metric deciles with higher scores inferring less biologically altered (i.e., better) conditions. Species distribution models were created for key sensitive and gamefish species (including Brook Trout, Northern Hog Sucker, Smallmouth Bass, and Torrent Sucker) to predict species occurrence. Uncertainty was calculated for both approaches using model prediction intervals. For complete data descriptions and data interpretation see associated publication (Maloney et al., 2022; https://doi.org/10.1016/j.ecolind.2021.108488).