This dataset contains site information, basin characteristics, results of flood-frequency analysis, and results of Bayesian weighted least-squares/Bayesian generalized least-squares (B-WLS/B-GLS) analysis of regional skewness of the annual peak flows for 785 streamflow gaging stations (streamgages) operated by the U.S. Geological Survey (USGS) in the Tennessee and parts of the Ohio and Lower Mississippi River basins (hydrologic unit codes 06, 05, and 08, respectively) in Tennessee, Kentucky, western Virginia, western West Virginia, far western Maryland and parts of North Carolina, Georgia, Alabama, and Mississippi. Annual peak-flow data through the 2021 water year (a water year is defined as the period October 1-September 30 and named for the year in which it ends) were used in the study. For regional skew analysis, 283 of the 785 candidate streamgages were removed for pseudo record length (PRL; Veilleux and Wagner, 2021) less than 30 years, 108 were removed for redundancy, 4 were removed for regulation and 2 were removed for urbanization (see file "VAskew_Region2.csv" in this dataset). For the remaining 387 of 785 candidate streamgages, B-WLS/B-GLS regression (Veilleux and Wagner, 2021) was used to relate flood skew to a suite of 32 explanatory variables. None of the explanatory variables tested had sufficient predictive power in explaining the variability in skew in the region; thus, a constant model of regional skew, 0.048 (average variance of prediction 0.16, standard error 0.4) was selected for the study area (Messinger and others, 2025). For the 785 candidate streamgages, annual peak-flow data through the 2021 water year ("VAskew_region2.pkf") and specification ("VAskew_region2.psf"), output ("VASKEW_REGION2.PRT"), and export ("VASKEW_REGION2.EXP") files from flood-frequency analysis in version 7.4.1 of USGS PeakFQ software (hereafter referred to as "PeakFQ"; Veilleux and others, 2014; Flynn and others, 2006) are provided. Two .csv files are provided, one describing the basin characteristics tested ("BasinCharsTested.csv") and the other ("VAskew_Region2.csv") containing site information (U.S. Geological Survey, 2023), results of flood-frequency analysis in PeakFQ, and, for the 387 streamgages used in the B-WLS/B-GLS regression, PRL, unbiased at-site skew, unbiased mean squared error of the at-site skew, the B-WLS/B-GLS residual, and metrics of leverage and influence. A geographic information systems (GIS) shapefile ("VA_SkewRegion2.shp") containing a polygon representing the geographic extent of the skew region is also included.