Meteorologic and hydrologic data from five small watersheds in the coal areas of West Virginia were used to calibrate and test the U.S. Geological Survey Precipitation-Runoff Modeling System for simulating streamflow under various climatic and land-use conditions. Three of the basins--Horsecamp Run, Gilmer Run, and Collison Creek--are primarily forested and relatively undisturbed. The remaining basins--Drawdy Creek and Brier Creek-are extensively mined, both surface and underground above stream drainage level. Low-flow measurements at numerous synoptic sites in the mined basins indicate that coal mining has substantially altered the hydrologic system of each basin. The effects of mining on streamflow that were identified are (1) reduced base flow in stream segments underlain by underground mines, (2) increased base flow in streams that are downdip and stratigraphically below the elevation of the mined coal beds, and (3) interbasin transfer of ground water through underground mines. These changes probably reflect increased permeability of surface rocks caused by subsidence fractures associated with collapsed underground mines in the basin. Such fractures would increase downward percolation of precipitation, surface and subsurface flow, and ground-water flow to deeper rocks or to underground mine workings. Model simulations of the water budgets for the unmined basins during the 1972-73 water years indicate that total annual runoff averaged 60 percent of average annual precipitation; annual evapotranspiration losses averaged 40 percent of average annual precipitation. Of the total annual runoff, approximately 91 percent was surface and subsurface runoff and 9 percent was groundwater discharge. Changes in storage in the soil zone and in the subsurface and ground-water reservoirs in the basins were negligible. In contrast, water-budget simulations for the mined basins indicate significant differences in annual recharge and in total annual runoff. Model simulations of the water budget for Drawdy Creek basin indicate that total annual runoff during 1972-73 averaged only 43 percent of average annual precipitation--the lowest of all study basins; annual evapotranspiration losses averaged 49 percent, and interbasin transfer of ground-water losses averaged about 8 percent. Of the total annual runoff, approximately 74 percent was surface and subsurface flow and 26 percent was ground-water discharge. The low total annual runoff at Drawdy Creek probably reflects increased recharge of precipitation and surface and subsurface flow losses to ground water. Most of the increase in ground-water storage is, in turn, lost to a ground-water sink--namely, interbasin transfer of ground water by gravity drainage and (or) mine pumpage from underground mines that extend to adjacent basins. Hypothetical mining situations were posed for model analysis to determine the effects of increased mining on streamflow in the mined basins. Results of model simulations indicate that streamflow characteristics, the water budget, and the seasonal distribution of streamflow would be significantly modified in response to an increase in mining in the basins. Simulations indicate that (1) total annual runoff in the basins would decrease because of increased surface- and subsurface-flow losses and increased recharge of precipitation to ground water (these losses would tend to reduce medium to high flows mainly during winter and spring when losses would be greatest), (2) extreme high flows in response to intense rainstorms would be negligibly affected, regardless of the magnitude of mining in the basins, (3) ground-water discharge also would decrease during winter and spring, but the amount and duration of low flows during summer and fall would substantially increase in response to increased ground-water storage in rocks and in underground mines, and (4) the increase in ground-water storage in the basins would be depleted, mostly by increased losses to a grou