The Leetown Science Center is a research facility operated by the U.S. Geological Survey that occupies approximately 455-acres near Kearneysville, Jefferson County, West Virginia. Aquatic and fish research conducted at the Center requires adequate supplies of high-quality, cold ground water. Three large springs and three production wells currently (in 2006) supply water to the Center. The recent construction of a second research facility (National Center for Cool and Cold Water Aquaculture) operated by the U.S. Department of Agriculture and co-located on Center property has placed additional demands on available water resources in the area. A three-dimensional steady-state finite-difference ground-water flow model was developed to simulate ground-water flow in the Leetown area and was used to assess the availability of ground water to sustain current and anticipated future demands. The model also was developed to test a conceptual model of ground-water flow in the complex karst aquifer system in the Leetown area. Due to the complexity of the karst aquifer system, a multidisciplinary research study was required to define the hydrogeologic setting. Geologic mapping, surface- and borehole-geophysical surveys, stream base-flow surveys, and aquifer tests were conducted to provide the hydrogeologic data necessary to develop and calibrate the model. It would not have been possible to develop a numerical model of the study area without the intensive data collection and methods developments components of the larger, more comprehensive hydrogeologic investigation. Results of geologic mapping and surface-geophysical surveys verified the presence of several prominent thrust faults and identified additional faults and other complex geologic structures (including overturned anticlines and synclines) in the area. These geologic structures are known to control ground-water flow in the region. Results of this study indicate that cross-strike faults and fracture zones are major avenues of ground-water flow. Prior to this investigation, the conceptual model of ground-water flow for the region focused primarily on bedding planes and strike-parallel faults and joints as controls on ground-water flow but did not recognize the importance of cross-strike faults and fracture zones that allow ground water to flow downgradient across or through less permeable geologic formations. Results of the ground-water flow simulation indicate that current operations at the Center do not substantially affect either streamflow (less than a 5-percent reduction in annual streamflow) or ground-water levels in the Leetown area under normal climatic conditions but potentially could have greater effects on streamflow during long-term drought (reduction in streamflow of approximately 14 percent). On the basis of simulation results, ground-water withdrawals based on the anticipated need for an additional 150 to 200 gal/min (gallons per minute) of water at the Center also would not seriously affect streamflow (less than 8 to 9 percent reduction in streamflow) or ground-water levels in the area during normal climatic conditions. During drought conditions, however, the effects of current ground-water withdrawals and anticipated additional withdrawals of 150 to 200 gal/min to augment existing supplies result in moderate to substantial declines in water levels of 0.5-1.2 feet (ft) in the vicinity of the Center's springs and production wells. Streamflow was predicted to be reduced locally by approximately 21 percent. Such withdrawals during a drought or prolonged period of below normal ground-water levels would result in substantial declines in the flow of the Center's springs and likely would not be sustainable for more than a few months. The drought simulated in this model was roughly equivalent to the more than 1-year drought that affected the region from November 1998 through February 2000. The potential reduction in streamflow is a result of capture of ground water tha