Temperatures of 100??C are measured at 3 km depth in a well located on the resurgent dome in the center of Long Valley Caldera, California, despite an assumed >800??C magma chamber at 6-8 km depth. Local downflow of cold meteoric water as a process for cooling the resurgent dome is ruled out by a Pecle??t-number analysis of temperature logs. These analyses reveal zones with fluid circulation at the upper and lower boundaries of the Bishop Tuff, and an upflow zone in the metasedimentary rocks. Vertical Darcy velocities range from 10 to 70 cm a-1. A 21-km-long geoelectrical profile across the caldera provides resistivity values to the order of 100 to >103 ??m down to a depth of 6 km, as well as variations of self-potential. Interpretation of the electrical data with respect to hydrothermal fluid movement confirms that there is no downflow beneath the resurgent dome. To explain the unexpectedly low temperatures in the resurgent dome, we challenge the common view that the caldera as a whole is a regime of high temperatures and the resurgent dome is a local cold anomaly. Instead, we suggest that the caldera was cooled to normal thermal conditions by vigorous hydrothermal activity in the past, and that a present-day hot water flow system is responsible for local hot anomalies, such as Hot Creek and the area of the Casa Diablo geothermal power plant. The source of hot water has been associated with recent shallow intrusions into the West Moat. The focus of planning for future power plants should be to locate this present-day flow system instead of relying on heat from the old magma chamber. ?? 2003 Elsevier B.V. All rights reserved.