The State of Idaho and local water users are concerned that streamflow depletion in the Portneuf River in Caribou and Bannock Counties is linked to ground-water withdrawals for irrigated agriculture. A year-long field study during 2001 02 that focused on monitoring surface- and ground-water relations was conducted, in cooperation with the Idaho Department of Water Resources, to address some of the water-user concerns. The study area comprised a 10.2-mile reach of the Portneuf River downstream from the Chesterfield Reservoir in the broad Portneuf Valley (Portneuf River Valley reach) and a 20-mile reach of the Portneuf River in a narrow valley downstream from the Portneuf Valley (Pebble-Topaz reach). During the field study, the surface- and ground-water relations were dynamic. A losing river reach was delineated in the middle of the Portneuf River Valley reach, centered approximately 7.2 miles downstream from Chesterfield Reservoir. Two seepage studies conducted in the Portneuf Valley during regulated high flows showed that the length of the losing river reach increased from 2.6 to nearly 6 miles as the irrigation season progressed.Surface- and ground-water relations in the Portneuf Valley also were characterized from an analysis of specific conductance and temperature measurements. In a gaining reach, stratification of specific conductance and temperature across the channel of the Portneuf River was an indicator of ground water seeping into the river.An evolving method of using heat as a tracer to monitor surface- and ground-water relations was successfully conducted with thermistor arrays at four locations. Heat tracing monitored a gaining reach, where ground water was seeping into the river, and monitored a losing reach, where surface water was seeping down through the riverbed (also referred to as a conveyance loss), at two locations.Conveyance losses in the Portneuf River Valley reach were greatest, about 20 cubic feet per second, during the mid-summer regulated high flows. Conveyance losses in the Pebble-Topaz reach were greatest, about 283 cubic feet per second, during the spring regulated high flows and were attributed to a hydroelectric project.Comparison of water levels in 30 wells in the Portneuf Valley during September and October 1968 and 2001 indicated long-term declines since 1968; the median decline was 3.4 feet. September and October were selected for characterizing long-term ground-water-level fluctuations because declines associated with irrigation reach a maximum at the end of the irrigation season. The average annual snowpack in the study area has declined significantly; 1945 85 average annual snowpack was 16.1 inches, whereas 1986 through 2002 average annual snowpack was 11.6 inches. Water-level declines during 1998  2002 may be partially attributable to the extended dry climatic conditions. It is unclear whether the declines could be partially attributed to increases in ground-water withdrawals. Between 1968 and 1980, water rights for ground-water withdrawals nearly doubled from 23,500 to 46,000 acre-feet per year. During this period, ground-water levels were relatively constant and did not exhibit a declining trend that could be related to increased ground-water withdrawal rights. However, ground-water withdrawals are not measured in the valley; thus, the amount of water pumped is not known. Since the 1990s, there have been several years when the Chesterfield Reservoir has not completely refilled, and the water in storage behind the reservoir has been depleted by the middle of the irrigation season. In this situation, surface-water diversions for irrigation were terminated before the end of the irrigation season, and irrigators, who were relying in part on diversions from the Portneuf River, had to rely solely on ground water as an alternate supply. Smaller volumes of water in the Chesterfield Reservoir since the 1990s indicate a growing demand for ground-water supplies.