The U.S. Air Force operates a major aerospace systems testing facility at Arnold Engineering Development Center (AEDC) in Coffee County, Tennessee. Dewatering operations at one of the test facilities, the J4 test cell, has affected the local ground-water hydrology. The J4 test cell is approximately 100 feet in diameter, extends approximately 250 feet below land surface, and penetrates several aquifers. Ground water is pumped continuously from around the test cell to keep the cell structurally intact. Because of the test cell's depth, dewatering has depressed water levels in the aquifers surrounding the site. The depressions that have developed exhibit anisotropy that is controlled by zones of high permeability in the aquifers. Additionally, contaminants - predominately volatile organic compounds - are present in the ground-water discharge from the test cell and in ground water at several other Installation Restoration Program (IRP) sites within the AEDC facility. The dewatering activities at J4 are drawing these contaminants from the nearby sites. The effects of dewatering at the J4 test cell were investigated by studying the lithologic and hydraulic characteristics of the aquifers, investigating the anisotropy and zones of secondary permeability using geophysical techniques, mapping the potentiometric surfaces of the underlying aquifers, and developing a conceptual model of the ground-water-flow system local to the test cell. Contour maps of the potentiometric surfaces in the shallow, Manchester, and Fort Payne aquifers (collectively, part of the Highland Rim aquifer system) show anisotropic water-level depressions centered on the J4 test cell. This anisotropy is the result of features of high permeability such as chert-gravel zones in the regolith and fractures, joints, and bedding planes in the bedrock. The presence of these features of high permeability in the Manchester aquifer results in complex flow patterns in the Highland Rim aquifers near the J4 test cell. The occurrence, distribution, and orientation of these features has a great effect on ground-water flow to the J4 test cell. The depression caused by dewatering extends out horizontally through the aquifers along the most permeable pathways. Since the aquifers above the Chattanooga Shale are not separated by distinct confining units, areas in adjacent aquifers above and below these zones of high permeability in the Manchester aquifer are also dewatered. Conditions in all Highland Rim aquifers approximate steady-state equilibrium because ground-water withdrawal at the test cell has been continuous since the late 1960's. The average ground-water discharge from the dewatering system at the J4 test cell was 105 gallons per minute, for 1992-95. The ground-water capture areas in each aquifer extend into all or parts of landfill #2 and leaching pit #2 (IRP site 1), the main testing area (IRP site 7), and the old fire training area (IRP site 10). IRP sites 8 and 12 are outside the ground-water capture areas. Of the 35 sampled wells in the J4 area, 10 produced water samples containing chlorinated organic compounds such as 1,2-dichloroethane (1,2-DCA), 1,1-dichloroethylene (1,1-DCE), and trichloroethylene (TCE) in concentrations which exceeded the Tennessee Department of Environment and Conservation Maximum Contaminant Levels (MCL's) for public water-supply systems. The highest concentrations were detected in samples from well AEDC-274 with 45 micrograms per liter (mg/L) 1,2-DCA, 320 mg/L 1,1-DCE, and 1,200 mg/L TCE. These compounds are synthetic and do not occur naturally in the environment. A sample of the ground-water discharge from the J4 test cell also contained concentrations of these compounds that exceed MCL's. Chlorinated organic compounds, including 1,2-DCA; 1,1-DCE; and TCE also have been detected at IRP sites 1, 7, 8, nd 10. The six dewatering wells surrounding the J4 test cell penetrate the Chattanooga Shale and are open to the Highland Rim aquifer system, there