The Pierina high-sulfidation Au-Ag deposit formed 14.5 my ago in rhyolite ash flow tuffs that overlie porphyritic andesite and dacite lavas and are adjacent to a crosscutting and interfingering dacite flow dome complex. The distribution of alteration zones indicates that fluid flow in the lavas was largely confined to structures but was dispersed laterally in the tuffs because of a high primary and alteration-induced permeability. The lithologically controlled hydrodynamics created unusual fluid, temperature, and pH conditions that led to complete SO42--H2S isotopic equilibration during the formation of some magmatic-steam and steam-heated alunite, a phenomenon not previously recognized in similar deposits. Isotopic data for early magmatic hydrothermal and main-stage alunite (??34S=8.5??? to 31.7???; ??18 OSO4=4.9??? to 16.5???; ??18 OOH=2.2??? to 14.4???; ??D=-97??? to -39???), sulfides (??34 S=-3.0??? to 4.3???), sulfur (??34S=-1.0??? to 1.1???), and clay minerals (??18O=4.3??? to 12.5???; ??D=-126??? to -81???) are typical of high-sulfidation epithermal deposits. The data imply the following genetic elements for Pierina alteration-mineralization: (1) fluid and vapor exsolution from an I-type magma, (2) wallrock buffering and cooling of slowing rising vapors to generate a reduced (H2S/SO4???6) highly acidic condensate that mixed with meteoric water but retained a magmatic ??34S???S signature of ???1???, (3) SO2 disproportionation to HSO4- and H2S between 320 and 180 ??C, and (4) progressive neutralization of laterally migrating acid fluids to form a vuggy quartz???alunite-quartz??clay???intermediate argillic???propylitic alteration zoning. Magmatic-steam alunite has higher ??34S (8.5??? to 23.2???) and generally lower ??18OSO4 (1.0 to 11.5???), ??18OOH (-3.4 to 5.9???), and ??D (-93 to -77???) values than predicted on the basis of data from similar occurrences. These data and supporting fluid-inclusion gas chemistry imply that the rate of vapor ascent for this environment was unusually slow, which provided sufficient time for the uptake of groundwater and partial to complete SO42--H2S isotopic exchange. The slow steam velocities were likely related to the dispersal of the steam column as it entered the tuffs and possibly to intermediate exsolution rates from magmatic brine. The low ??D values may also partly reflect continuous degassing of the mineralizing magma. Similarly, data for steam-heated alunite (??34S=12.3??? to 27.2???; ??18OSO4=11.7??? to 13.0???; ??18OOH=6.6??? to 9.4???; ??D=-59??? to -42???) are unusual and indicate a strong magmatic influence, relatively high temperatures (140 to 180 ??C, based on ??18 OSO4-OH fractionations), and partial to complete sulfur isotopic exchange between steam-heated sulfate and H2S. Restricted lithologically controlled fluid flow in the host tuffs allowed magmatic condensate to supplant meteoric groundwater at the water table and create the high-temperature low-pH conditions that permitted unusually rapid SO42--H2S isotopic equilibration (50-300 days) and (or) long sulfate residence times for this environment. Late void-filling barite (??34S=7.4??? to 29.7???; ??18OSO4=-0.4??? to 15.1???) and later void-filling goethite (??18O=-11.8??? to 0.2???) document a transition from magmatic condensate to dominantly meteoric water in steam-heated fluids during cooling and collapse of the hydrothermal system. These steam-heated fluids oxidized the top ???300 m of the deposit by leaching sulfides, redistributing metals, and precipitating barite??acanthite??gold and goethite-hematite ??gold. Steam-heated oxidation, rather than weathering, was critical to forming the orebody in that it not only released encapsulated gold but likely enriched the deposit to ore-grade Au concentrations. ?? 2004 Elsevier B.V. All rights reserved.