A high-resolution geologic model of the General Separations Area supporting groundwater cleanup at the Savannah River Site (SRS) was developed using geostatistical approaches adapted from the petroleum industry. Depositional facies in each formation were identified from core descriptions and modeled using sequential indicator simulation. Petrophysical properties in each facies were then modeled as distinct fields using sequential Gaussian simulation. A complete property distribution was assembled as a montage of the three-dimensional fields simulated independently in each facies. Petrophysical property realizations thus reflect the spatial distribution of and sharp boundaries between facies. Application of sequence-stratigraphic principles was difficult because of the small scale of the study site, unconsolidated nature of sediments, and proximity to the Fall Line. Nevertheless, multidimensional variogram mapping was effective at identifying the underlying spatial structures in the field data needed to reproduce major preserved features of the ancient coastline in model surfaces and facies and property fields. Preliminary assessment of permeability fields derived from the geologic model using connectivity and streamline and tracer simulation suggests that the methods were successful in identifying preferential pathways for contaminant migration from the SRS F-area seepage basins. However, further analysis using a numerical solute transport model with rigorous boundary conditions is needed for confirmation.