Expanding on current efforts to evaluate the role of groundwater dynamics in managing and restoring Chesapeake Bay (USA), the integrated hydrologic model ParFlow-CLM was applied to a 374,976-km2 area encompassing the Chesapeake Bay watershed. The model included a representation of surface water, groundwater and land-surface energy fluxes with spatially variable atmospheric forcing at an hourly time step. The study tackled issues of data availability, access, assembly, and synthesis for estimating hydrogeologic properties in the context of the development of a large-scale model. Hydrogeologic properties from literature and other sources were assembled, processed, and synthesized to derive a conceptual hydrogeologic model consisting of 29 hydrofacies and a three-dimensional hydraulic conductivity field. Evaluation of the ParFlow-CLM model output showed that the constructed model captured seasonal and spatial variability in subsurface storage, surface storage and surface runoff, and produced water-table depths consistent with the topography, meteorological forcing, and hydrogeology. Comparison with well data from the US Geological Survey showed good agreement of model output with observed hydraulic heads for most of the data. Modeled terrestrial water storage changes compared well with GRACE satellite data with a root mean square error of 2.3 cm. Model results showed the dominant contribution of subsurface storage changes (90%) to terrestrial water storage changes in the region.