Carlin-type deposits are major sources of gold, yet their origins are enigmatic. Suggested genetic models make connections to magmatism, regional metamorphism, or regional extension. Depositional mechanisms are uncertain as well. We propose on the basis of geologic, physical, and chemical reasoning, a genetic model in which meteoric fluids were circulated by heat released during crustal extension. These fluids interacted at depth with the sedimentary rock pile and scavenged gold. Upon upwelling, these fluids interacted with various lithologies and/or other fluids and produced the characteristic alteration and metal suites of these deposits. To test the viability of this amagmatic model, we have investigated certain physical and chemical constraints implicit to the model. Heat balance calculations indicate that ample surface waters could be heated to appropriate temperatures and circulated during rapid crustal extension. Mass balance calculations indicate that solution transport efficiencies of <10 percent can account for several times the amount of known alteration and mineralization. In contrast, magma-driven or metamorphic fluids appear to be less adequate sources. A two-step, source to deposit approach was used to simulate fluid-rock interaction and mass transport implied in the amagmatic model. Fluids with a fixed Sigma Cl were first equilibrated with various potential source rocks (arkose, graywacke, and pelite, each with various f O2 and f S2 buffer assemblages) at 300 degrees C and 1 kbar to simulate reaction at depth, perhaps near the brittle-ductile transition. These fluids were then reequilibrated under conditions indicated for deposit formation (225 degrees -150 degrees C, 0.5 kbars) to investigate processes including cooling, wall-rock reaction, and mixing which could be important to ore deposition. Gold transport was favored for low-chlorinity, intermediate oxidation state (below pyrrhotite-pyrite-magnetite), slightly acid fluids from sulfide-bearing sources (i.e., equilibration with arkosic rocks). These conditions are not conducive to base metal transport and explain the rarity of base metals in Carlin-type systems. When reacted under deposit level conditions, these calculated ore fluids leach carbonate and deposit quartz. The volume ratio of carbonate dissolved to quartz precipitated is approximately 0.2 to 2.5, and increases when mixing with a second fluid is considered. Arkose-equilibrated source fluids can precipitate gold by combinations of wall-rock reaction with a decarbonated host, mixing with various sulfide-poor fluids, and reaction with iron-rich host rocks. Cooling alone is ineffective over this temperature interval. Although most deposits are hosted by decarbonated marls, geochemical calculations suggest that other depositional environments are feasible, consistent with observation. Although here applied to the amagmatic hypothesis, these chemical calculations are not mechanism specific and might also be appropriate to metamorphic or magma-driven hydrothermal systems. These results demonstrate that Carlin-type deposits can result from the coincidence of an effective driving force (rapid, major crustal extension), a source of dilute fluids, and chemically appropriate sets of both source and host rocks. The aqueous chemistry involved is unexceptional and shares many features with other types of moderate- to low-temperature gold-bearing hydrothermal systems--it is the geologic setting, and perhaps the fluid driving force, that is unusual. An amagmatic model is consistent with both deposit-scale and regional-scale observations; it rationalizes Carlin-type systems in the framework of southwestern North American metallogeny and it implies multiple possible mechanisms for gold mineralization. Detachment-type gold systems may be the high-oxidation state, high-salinity analogues of Carlin-type systems, formed by major extension, but from contrasting lithologic and fluid sources.