We present a thermodynamic model for mineral dissolution in aqueous fluids at elevated temperatures and pressures, based on intrinsic thermal properties and variations of volumetric properties of the aqueous solvent. The standard thermodynamic properties of mineral dissolution into aqueous fluid consist of two contributions: one from the energy of transformation from the solid to the hydrated-species state and the other from the compression of solvent molecules during the formation of a hydration shell. The latter contribution has the dimension of the generalized Krichevskii parameter. This approach describes the energetics of solvation more accurately than does the Born electrostatic theory and can be extended beyond the limits of experimental measurements of the dielectric constant of H 2 O. The new model has been calibrated by experimental solubilities of quartz, corundum, rutile, calcite, apatite, fluorite and portlandite in pure H 2 O at temperatures up to 1100°C and pressures up to 20 kbar. All minerals show a steady increase in solubility along constant geothermal gradients or water isochores. By contrast, isobaric solubilities initially increase with rising temperature but then decline above 200-400°C. This retrograde behavior is caused by variations in the isobaric expansivity of the aqueous solvent, which approaches infinity at its critical point. Oxide minerals predominantly dissolve to neutral species; so, their dissolution energetics involve a relatively small contribution from the solvent volumetric properties and their retrograde solubilities are restricted to a relatively narrow window of temperature and pressure near the critical point of water. By contrast, Ca-bearing minerals dissolve to a variety of charged species; so, the energetics of their dissolution reactions involve a comparatively large contribution from volume changes of the aqueous solvent and their isobaric retrograde solubility spans nearly all metamorphic and magmatic conditions. These features correlate with and can be predicted from the standard partial molar volumes of aqueous species.The thermodynamic model can be used over much wider range of settings for terrestrial fluid-rock interaction than has previously been possible. To illustrate, it is integrated with transport theory to show quantitatively that integrated fluid fluxes characteristic of crustal shear zones are capable of precipitating quartz or calcite veins from low-and medium-grade metamorphic conditions, at a geothermal gradient of 20°C km )1 . For subduction zones, modeled by a geotherm of 7°C km )1 , the required fluid fluxes are one to two orders of magnitude lower and predict enhanced efficiency of mass transfer and metasomatic precipitation in comparison with orogenic settings. The new model thus can be applied to shallow hydrothermal, metamorphic, magmatic and subduction fluids, and for retrieval of dependent thermodynamic properties for mass transfer or geodynamic modeling.