Two small watersheds in eastern Ohio that were surface mined for coal and reclaimed were studied during 1986-89. Water-level and water-quality data were compared with similar data collected during previous investigations conducted during 1976-83 to determine long-term effects of surface mining on the hydrologic system. Before mining, the watersheds were characterized by sequences of flat-lying sedimentary rocks containing two major coal seams and underclays. An aquifer was present above each of the underclays. Surface mining removed the upper aquifer, stripped the coal seam, and replaced the sediment. This created a new upper aquifer with different hydraulic and chemical characteristics. Mining did not disturb the middle aquifer. A third, deeper aquifer in each watershed was not studied. Water levels were continuously recorded in one well in each aquifer. Other wells were measured every 2 months. Water levels in the upper aquifers reached hydraulic equilibrium from 2 to 5 years after mining ceased. Water levels in the middle aquifers increased more than 5 feet during mining and reached equilibrium almost immediately thereafter. Water samples were collected from three upper-aquifer well, a seep from the upper aquifer, and the stream in each watershed. Two samples were collected in 1986 and 1987, and one each in 1988 and 1989. In both watersheds, sulfate replaced bicarbonate as the dominant upper-aquifer and surface-water anion after mining. For the upper aquifer of a watershed located in Muskingum County, water-quality data were grouped into premining and late postmining time periods (1986-89). The premining median pH and concentration of dissolved solids and sulfate were 7.6, 378 mg/L (milligrams per liter), and 41 mg/L, respectively. The premining median concentrations of iron and manganese were 10? /L (micrograms per liter) and 25?, respectively. The postmining median values of pH, dissolved solids, and sulfate were 6.7, 1,150 mg/L, and 560 mg/L, respectively. The postmining median concentrations of iron and manganese were 3,900?g/L and 1,900? g/L, respectively. For the upper aquifer of a watershed located in Jefferson County, the water-quality data were grouped into three time periods of premining, early postmining, and late postmining. The premining median pH and concentrations of dissolved solids and sulfate were 7.0, 335 mg/L, and 85 mg/L, respectively. The premining median concentrations of iron and manganese were 30? g/L for each constituent. Late postmining median pH and concentrations of dissolved solids and sulfate were 6.7, 1,495 mg/L, and 825 mg/L, respectively. The postmining median concentrations of iron and manganese were 31? g/L and 1,015? g/L, respectively. Chemistry of water in the middle aquifer in each watershed underwent similar changes. In general, statistically significant increases in concentrations of dissolved constituents occurred because of surface mining. In some constituents, concentrations increased by more than an order of magnitude. The continued decrease in pH indicated that ground water had no reached geochemical equilibrium in either watershed more than 8 years after mining.