A natural-gradient ground-water tracer test was designed and conducted in a tidal freshwater wetland at West Branch Canal Creek, Aberdeen Proving Ground, Maryland. The objectives of the test were to characterize solute transport at the site, obtain data to more accurately determine the ground-water velocity in the upper wetland sediments, and to compare a conservative, ionic tracer (bromide) to a volatile tracer (sulfur hexafluoride) to ascertain whether volatilization could be an important process in attenuating volatile organic compounds in the ground water. The tracer test was conducted within the upper peat unit of a layer of wetland sediments that also includes a lower clayey unit; the combined layer overlies an aquifer. The area selected for the test was thought to have an above-average rate of ground-water discharge based on ground-water head distributions and near-surface detections of volatile organic compounds measured in previous studies. Because ground-water velocities in the wetland sediments were expected to be slow compared to the underlying aquifer, the test was designed to be conducted on a small scale. Ninety-seven ?-inch-diameter inverted-screen stainless-steel piezometers were installed in a cylindrical array within approximately 25 cubic feet (2.3 cubic meters) of wetland sediments, in an area with a vertically upward hydraulic gradient. Fluorescein dye was used to qualitatively evaluate the hydrologic integrity of the tracer array before the start of the tracer test, including verifying the absence of hydraulic short-circuiting due to nonnatural vertical conduits potentially created during piezometer installation. Bromide and sulfur hexafluoride tracers (0.139 liter of solution containing 100,000 milligrams per liter of bromide ion and 23.3 milligrams per liter of sulfur hexafluoride) were co-injected and monitored to generate a dataset that could be used to evaluate solute transport in three dimensions. Piezometers were sampled 2 to 15 times each, from July 1998 through September 1999, to assess background conditions and monitor tracer movement. During the test, 644 samples were analyzed for fluorescein, 617 samples were analyzed for bromide with an ion-selective electrode, 213 samples were analyzed for bromide with colorimetric methods, and 603 samples were analyzed for sulfur hexafluoride, including samples collected prior to tracer injection to determine background concentrations. Additional samples were analyzed for volatile organic compounds (96 samples) and methane (37 samples) to determine the distribution of these contaminants and the extent of methanogenic conditions within the tracer array; however, these data were not used for the analysis of the test. During the tracer test, the fluorescein dye, bromide, and sulfur hexafluoride were transported predominantly in the upward direction, although all three tracers also moved outward in all directions from the injection point, and it is likely that some tracer mass moved beyond the lateral edges of the array. An analysis of the tracer-test data was performed through the use of breakthrough curves and isoconcentration contour plots. Results show that movement of the fluorescein dye, a non-conservative tracer, was retarded compared to the other two tracers, likely as a result of sorption onto the wetland sediments. Suspected loss of tracer mass along the lateral edges of the array prevented a straightforward quantitative analysis of tracer transport and ground-water velocity from the bromide and sulfur-hexafluoride data. In addition, the initial density of the bromide/sulfur hexafluoride solution (calculated to be 1.097 grams per milli2 Ground-Water Tracer Test, West Branch Canal Creek, Aberdeen Proving Ground, MD liter) could have caused the solution to sink below the injection point before undergoing dilution and moving back up into the array. For these reasons, the data analysis in this report was performed largely through qualitative method