The U.S. Geological Survey, in cooperation with the Susquehanna River Basin Commission and the Pennsylvania Department of Environmental Resources, investigated the effects of agricultural best-management practices on surface-water quality as part of the U.S. Environmental Protection Agency's Chesapeake Bay Program. This report characterizes a 0.63-square- mile agricultural watershed underlain by shale, mudstone, and red arkosic sandstone in the Lower Susquehanna River Basin. The water quality of the Brush Run Creek site was studied from October 1985 through September 1991, prior to and during the implementation of nutrient management designed to reduce sediment and nutrient discharges into Conewago Creek, a tributary to the Chesapeake Bay. The original study area was 0.38 square mile and included an area immediately upstream from a manure lagoon. The study area was increased to 0.63 square mile in the fall of 1987 after an extensive tile-drain network was discovered upstream and downstream from the established streamflow gage, and the farm owner made plans to spray irrigate manure to the downstream fields. Land use for about 64 percent of the 0.63 square mile watershed is cropland, 14 percent is pasture, 7 percent is forest, and the remaining 15 percent is yards, buildings, water, or gardens. About 73 percent of the cropland was used to produce corn during the study. The average annual animal population consisted of 57,000 chickens, 1,530 hogs, and 15 sheep during the study. About 59,340 pounds of nitrogen and 13,710 pounds of phosphorus were applied as manure and commercial fertilizer to fields within the subbasin during the 3-year period prior to implementation of nutrient management. During nutrient management, about 14 percent less nitrogen and 57 percent less phosphorus were applied as commercial and manure fertilizer. Precipitation totaled 209 inches, or 13 percent less than the long-term normal, during the 6-year study. Concentrations of total ammonia in precipitation were as high as 2.7 mg/L (milligrams per liter); in dry deposition the concentrations were as high as 5.4 mg/L, probably because of the ammonia that had volatilized from the manure-storage lagoon. Nitrate nitrogen in the upper 4 feet of the soil ranged from 17 to 452 pounds per acre and soluble phosphorus content ranged from 0.29 to 65 pounds per acre. The maximum concentration of total nitrogen was 2,400 mg/L on September 10, 1986, in discharge from the tile drain near the streamflow gage. Median concentrations of total nitrogen and dissolved nitrite plus nitrate in base flow at the water-quality gage were 14 mg/L and 4.4 mg/L, respectively; prior to nutrient management and during nutrient management, median concentrations were 14 mg/L and 6.2 mg/L, respectively. Significant reductions in total phosphorus and suspended-sediment concentrations occurred at the water-quality gage. The maximum concentrations of total phosphorus (160 mg/L) and suspended sediment (3,530 mg/L) were measured at a tile line above the water-quality gage. Concentrations of total nitrogen, dissolved ammonia, and total phosphorus in base flow increased during dry periods when discharges from the tile drain were not diluted. During nutrient management, only base-flow loads of suspended sediment increased. Total streamflow was about 121.8 inches. About 81 percent was storm runoff. Loads of total nitrogen, total phosphorus in stormflow, and suspended sediment increased 14, 44, and 41 percent during nutrient management, respectively. A load of about 787,780 pounds of sediment, 22,418 pounds of nitrogen, and 5,479 pounds of phosphorus was measured during 214 sampled stormflow days that represented 84 percent of the stormflow. About 812,924 pounds of sediment, 38,421 pounds of nitrogen, and 6,377 pounds of phosphorus were discharged during the 6-year study.