A mechanistic model is presented to describe closure of a fracture mediated by pressure solution; closure controls permeability reduction and incorporates the serial processes of dissolution at contacting asperities, interfacial diffusion, and precipitation at the free face of fractures. These processes progress over a representative contacting asperity and define compaction at the macroscopic level, together with evolving changes in solute concentration for arbitrarily open or closed systems for prescribed ranges of driving effective stresses, equilibrium fluid and rock temperatures, and fluid flow rates. Measured fracture surface profiles are applied to define simple relations between fracture wall contact area ratio and fracture aperture that represents the irreversible alteration of the fracture surface geometry as compaction proceeds. Comparisons with experimental measurements of aperture reduction conducted on a natural fracture in novaculite <Polak et al., 2003> show good agreement if the unknown magnitude of microscopic asperity contact area is increased over the nominal fracture contact area. Predictions of silica concentration slightly underestimate the experimental results even for elevated microscopic contact areas and may result from the unaccounted contribution of free face dissolution. For the modest temperatures (20--150¿C) and short duration (900 hours) of the test, pressure solution is demonstrated to be the dominant mechanism contributing to both compaction and permeability reduction, despite net dissolution and removal of mineral mass. Pressure solution results in an 80% reduction in fracture aperture from 12 ¿m, in contrast to a ~10 nm contribution by precipitation, even for the case of a closed system. For the considered dissolution-dominated system, fracture closure rates are shown to scale roughly linearly with stress increase and exponentially with temperature increase, taking between days and decades for closure to reach completion.