The distribution, petrography, and mineralogy are described for hydrothermal eruption breccias in the active Waiotapu geothermal system, New Zealand. These surficial deposits are characterized by variable clast abundance, size, and angularity; a poorly sorted, matrix-supported texture; an absence of primary volcanic material; and a variable degree of hydrothermal alteration and brecciation between clasts and deposits. Clasts in some deposits provide evidence for transport from about a 300-m depth and eruption velocities of 200 to 300 m/s. Previously published discussions of the origin of these types of eruption, or equivalent breccias in epithermal mineralized districts, have focused on the development of substantial fluid overpressures in the upper silicified portion of a hydrothermal system. In this paper an alternative mechanism is proposed which requires only local sealing of near-surface discharge channels and the transmission of deeper reservoir pressures to the sealed area by the evolution of a compressible cap of exsolved gas (dominantly CO 2 ). Many of the Waiotapu eruption vents have been flooded by ground water, but a number, including the Champagne Pool which was formed 900 years ago, continue to discharge chloride water directly from the deeper hydrothermal reservoir. Precipitates formed within this pool are enriched in arsenic, antimony, thallium, and mercury and are ore grade with respect to gold and silver. Two chemical environments are recognized in and below such vents which (subsequent to the eruption) may lead to precious metal deposition: (1) boiling and H 2 S loss in the fracture network and conduit below the eruption crater and (2) coupled surface heat loss and pH buffering in the crater pool which leads to arsenic sulfide precipitation with coprecipitation of gold and other metals. It is argued that about one hundred thousand ounces of gold have deposited beneath Champagne Pool since its formation but that the overall transport and depositional efficiency of the Waiotapu system is only 10 percent. In 10,000 years this is sufficient for the deposition of 1- to 6-million-ounce orebodies of the type under exploitation at Round Mountain (Nevada), McLaughlin (California), and in other epithermal districts. Exploration programs for epithermal precious metal deposits should be based on our knowledge of the structural and chemical characteristics of active geothermal systems. Coupling patterns of trace element distribution, alteration assemblages, and fluid inclusion data with the location of hydrothermal breccias (vents and/or surficial deposits) provides a strategy for the targeting of exploration drilling by identifying flow paths along which boiling has occurred.