Regional aquifers underlying the 15,600-square-mile Snake River Plain in southern Idaho and eastern Oregon was studied as part of the U.S. Geological Survey's Regional Aquifer-System Analysis program. The largest and most productive aquifers in the Snake River Plain are composed of Quaternary basalt of the Snake River Group, which underlies most of the 10,8000-square-mile eastern plain. Aquifer tests and simulation indicate that transmissivity of the upper 200 feet of the basalt aquifer in the eastern plain commonly ranges from about 100,000 to 1,000,000 feet squared per day. However, transmissivity of the total aquifer thickness may be as much as 10 million feet squared per day. Specific yield of the upper 200 feet of the aquifer ranges from about 0.01 to 0.20. Average horizontal hydraulic conductivity of the upper 200 feet of the basalt aquifer ranges from less than 100 to 9,000 feet per day. Values may be one to several orders of magnitude higher in parts in individual flows, such as flow tops. Vertical hydraulic conductivity is probably several orders of magnitude lower than horizontal hydraulic conductivity and is generally related to the number of joints. Pillow lava in ancestral Snake River channels has the highest hydraulic conductivity of all rock types. Hydraulic conductivity of the basalt decreases with depth because of secondary filling of voids with calcite and silica. An estimated 80 to 120 million acre-feet of water is believed to be stored in the upper 200 feet of the basalt aquifer in the eastern plain. The most productive aquifers in the 4,800-square-mile western plain are alluvial sand and gravel in the Boise River valley. Although aquifer tests indicate that transmissivity of alluvium in the Boise River valley ranges from 5,000 to 160,000 feet squared per day, simulation suggests that average transmissivity of the upper 500 feet is generally less than 20,000 feet squared per day. Vertically averaged horizontal hydraulic conductivity of the upper 500 feet of alluvium ranges from about 4 to 40 feet per day; higher values can be expected in individual sand and gravel zones. Vertical hydraulic conductivity is considerably lower because of the presence of clay layers. Hydraulic heads measured in piezometers, interpreted from diagrams showing ground-water flow and equipotential lines and estimated by computer simulation, demonstrate that water movement is three dimensional through the rock framework. Natural recharge takes place along the margins of the plain where head decreases with depth; discharge takes place near some reaches of the Snake River and the Boise River where head increases with depth. Geothermal water in rhyolitic rocks in the western plain and western part of the eastern plain has higher hydraulic head than the overlying cold water. Geothermal water, therefore, moves upward and merges into the cold-water system. Basin water-budget analyses indicate that the volume of cold water. Carbon-14 age determinations, which indicate that residence time of geothermal water is 17,700 to 20,300 years, plus or minus 4,000 years, imply slow movement of water through the geothermal system. Along much of its length, the Snake River gains large quantities of ground water. On the eastern plain, the river gained about 1.9 million acre-feet of water between Blackfoot and Neeley, Idaho, in 1980. Between Milner and King Hill, Idaho, the river gained 4.7 million acre-feet, mostly as spring flow from the north side. Upstream from Blackfoot and in the vicinity of Lake Walcott, the rover loses flow to ground water during parts or all of the year. On the western plain, river gains from ground water are small relative to those on the eastern plain; most are from seepage. Streams in tributary drainage basins supply calcium/bicarbonate type and calcium/magnesium/bicarbonate type water to the plain. Water type is a reflection of the chemical composition of rocks in the drainage basin, Concentrations of dissolved solids are smallest, about 50 milligrams per liter, in streams such as the Boise River that drain areas of granitic rocks; concentrations are greatest, about 400 milligrams per liter, in streams such as the Owyhee and Raft Rivers that drain area of sedimentary rocks. Water chemistry reflects the interaction of surface water and ground water. The chemical composition of ground water in the plain is essentially the same as that in streamflow and groundwater discharge from tributary drainage basins. Tributary drainage basins supplied 85 percent of the ground-water recharge in the eastern plain during 1980 and a nearly equivalent percentage of the solute load in ground water; human activities and dissolution of minerals supplied the other solutes. Dissolved-solids concentrations in ground water were generally less than 400 milligrams per liter. Water from the lower geothermal system is chemically different from water from the upper cold-water system. Geothermal water typically has greater concentrations of sodium, bicarbonate, sulfate, chloride, fluoride, silica, arsenic, boron, and lithium and smaller concentrations of calcium, magnesium, and hydrogen. Difference are attributed to ion exchange as geothermal moves through the rock framework. Irrigation, mostly on the Snake River Plain, accounted for about 96 percent of consumptive water use in Idaho during 1980. The use of surface water for irrigation for more than 100 years has caused major changes in the hydrologic system on the plain. Construction of dams, reservoirs, and diversifications effected planned changes in the surface-water system but resulted in largely unplanned changes in the ground-water system. During those years of irrigation, annual recharge in the main part of the eastern plain increased to about 6.7 million acre-feet in 1980, or by about 70 percent. Most of the increase was from percolation of surface water diverted for irrigation. From preirrigation to 1952, groundwater storage increased about 24 million acre-feet, and storage decreased from 1952 to 1964 and from 1976 to 1980 because of below-normal precipitation and increased withdrawals of ground water for irrigation. Annual ground-water discharge increased to about 7.1 million acre-feet in 1980, or about 80 percent since the start of irrigation. About 10 percent of the 1980 total discharge was ground-water pumpage. About 3.1 million acres, or almost one-third of the plain, was irrigated during 1980: 2.0 million acres with surface water, 1.0 million acres with ground water, and 0.1 million acres with combined surface and ground water. About 8.9 million acre-feet of Snake River water was diverted for irrigation during 1980 and 2.3 million acre-feet of ground water was pumped from 5,300 wells. Most irrigation wells on the eastern plain are open to basalt. About two-thirds of them yield more than 1,500 gallons per minute with a reported maximum of 7,240 gallons per minute; drawdown is less than 20 feet in two-thirds of the wells. Most irrigation wells on the western plain are open to sedimentary rocks. About one-third of them yield more than 1,00 gallons per minute with a reported maximum of 3,850 gallons per minute; drawndown is less than 20 feet in about one-fifth of the wells. The major instream use of water on the Snake River Plain is hydroelectric power generation. Fifty-two million acre-feet of water generated 2.6 million megawatthours of electricity during 1980. Digital computer ground-water flows models of the eastern and western plain reasonably simulated regional changes in water levels and ground-water discharges from 1880 (preirrigation) to 1980. Model results support the concept of three-dimensional flow and the hypotheses of no underflow between the eastern and western plain. Simulation of the regional aquifer system in the eastern plain indicates that is 1980 hydrologic conditions, including pumpage, were to remain the same for another 30 years, moderate declines in ground-water levels and decreases in spring discharges would continue. Increased ground-water pumpage to irrigate an additional 1 million acres could cause ground-water levels to decline a few tens of feet in the central part of the plain and could cause corresponding decreases in ground-water discharge. A combination of actions such as increased ground-water pumpage and decreased use of surface water for irrigation (resulting in reduced recharge) would accentuate the changes.