An analog model of the ground-water basin of the upper Coachella Valley was constructed to determine the effects of imported water on ground-water levels. The model was considered verified when the ground-water levels generated by the model approximated the historical change in water levels of the ground-water basin caused by man's activities for the period 1986-67. The ground-water basin was almost unaffected by man's activities until about 1945 when ground-water development caused the water levels to begin to decline. The Palm Springs area has had the largest water-level decline, 75 feet since 1986, because of large pumpage, reduced natural inflow from the San Gorgonio Pass area, and diversions of natural inflows at Snow and Falls Creeks and Chino Canyon starting in 1945. The San Gorgonio Pass inflow had been reduced from about 18,000 acre-feet in 1986 to about 9,000 acre-feet by 1967 because of increased ground-water pumpage in the San Gorgonio Pass area, dewatering of the San Gorgonio Pass area that took place when the tunnel for the Metropolitan Water District of Southern California was drilled, and diversions of surface inflow at Snow and Falls Creeks. In addition, 1944-64 was a period of below-normal precipitation which, in part, contributed to the declines in water levels in the Coachella Valley. The Desert Hot Springs, Garnet Hill, and Mission Creek subbasins have had relatively little development; consequently, the water-level declines have been small, ranging from 5 to 15 feet since 1986. In the Point Happy area a decline of about 2 feet per year continued until 1949 when delivery of Colorado River water to the lower valley through the Coachella Canal was initiated. Since 1949 the water levels in the Point Happy area have been rising and by 1967 were above their 1986 levels. The Whitewater River subbasin includes the largest aquifer in the basin, having sustained ground-water pumpage of about 740,000 acre-feet from 1986 to 1967, and will probably continue to provide the most significant supply of ground water for the upper valley. The total ground-water storage depletion for the entire upper valley for 1986-67 was about 600,000 acre-feet, an average storage decrease of about 25,000 acre-feet per year since 1945. Transmissivity for the Whitewater River subbasin ranges from 860,000 gallons per day per foot (near Point Happy) to 50,000 gallons per day per foot, with most of the subbasin about 800,000 gallons per day per foot. In contrast, the transmissivities of the Desert Hot Springs, Mission Creek, and Garnet Hill subbasins generally range from 2,000 to 100,000, but the highest value, beneath the Mission Creek streambed deposits, is 200,000 gallons per day per foot; the transmissivity for most of the area of th6 three subbasins is 80,000 gallons per day per foot. The storage coefficients are representative of water-table conditions, ranging from 0.18 beneath the Mission Creek stream deposits to 0.06 in the Palm Springs area. The model indicated that the outflow at Point Happy decreased from 50,000 acre-feet in 1936 to 30,000 acre-feet by 1967 as a result of the rising water levels in the lower valley. The most logical area to recharge the Colorado River water is the Windy Point-Whitewater area, where adequate percolation rates of 2-4 acre-feet per acre per day are probable. The Whitewater River bed may be the best location to spread the water if the largest part of the imported water can be recharged during low-flow periods. The area in sec. 21, T. 2 S., R. 4 E., would be adequate for the smaller quantities of recharge proposed for the Mission Creek area. Projected pumpage for the period 1968-2000 was programmed on the model with the proposed recharge of Colorado River water for the same period. The model indicated a maximum water-level increase of 200 feet above the 1967 water level at Windy Point, the proposed recharge site, by the year 2000, a 130-foot increase by 1990, and a 20-foot increas