In the Central Valley of California, a great agricultural economy has been developed in a semiarid environment. This economy is supported by imported surface water and 9 to 15 million acre-feet per year of ground water. Estimates of ground-water pumpage computed from power consumption have been compiled and summarized. Under ideal conditions, the accuracy of the methods used is about 3 percent. This level of accuracy is not sustained over the entire study area. When pumpage for the entire area is mapped, the estimates seem to be consistent areally and through time. A multiple linear-regression model was used to synthesize data for the years 1961 through 1977, when power data were not available. The model used a relation between ground-water pumpage and climatic indexes to develop a full suite of pumpage data to be used as input to a digital ground-water model, one of the products of the Central Valley Aquifer Project. Statistical analysis of well-perforation data from drillers ' logs and water-temperature data was used to determine the percentage of pumpage that was withdrawn from each of two horizontal layers. (USGS)