Aquifer bulk resistivity and grain-surface resistivity (inverse of grain-surface conductance) were tested as geoelectrical analogs to the horizontal hydraulic conductivity of clastic, freshwater aquifers in the Southeastern United States. Bulk resistivity was also tested as a geoelectrical analog for dissolved-solids concentrations in aquifer water. Bulk resistivity was defined as the average resistivity across a contributing interval measured by the long-normal (64-inch) or induction log. Grain-surface resistivity was empirically defined as the difference between aquifer bulk resistivity and aquifer water resistivity (computed from specific conductance). Sources of data were borehole geophysical logs and results of water-quality and aquifer-test analyses related to unconsolidated sands and clayey sands at more than a hundred sites in seven Southeastern States. Waterbearing units were composed of sediments ranging from the Late Cretaceous to middle Eocene. All bivariate data were related using the logarithmic regression model Y=AX B. Aquifer bulk resistivity and grain-surface resistivity were moderately correlated to horizontal hydraulic conductivity (70 and 72 percent correlation coefficients, respectively). Apparent formation factor, defined as the ratio of aquifer bulk resistivity to aquifer water resistivity, was shown to be poorly correlated with horizontal hydraulic conductivity (38 percent correlation coefficient). Aquifer bulk resistivity was shown to be highly correlated with dissolved-solids concentration and aquifer water resistivity (88 and 93 percent correlation coefficients, respectively). Regression models using bulk resistivity and aquifer water resistivity as independent variables were applied at four locations in South Carolina and Louisiana to predict dissolved-solids concentrations in aquifer water. Absolute mean error of prediction was 20 and 6 percent, respectively. A regression model using bulk resistivity to predict horizontal hydraulic conductivity was applied at 27 sites in 6 Southeastern States, resulting in an absolute error ranging from 4 to 95 percent with a corresponding mean error of 43 percent.