The San Martin skarn deposit was formed by a hydrothermal system associated with intrusion of the 46-m.y.-old Cerro de la Gloria quartz monzonite stock into the middle Cretaceous Cuesta del Cura limestone. The deposit is exploited by two major mines. The San Martin mine extracts Cu-Zn-Ag ore from veins and replacement bodies hosted by skarn, and the Sabinas mine extracts Zn-Pb-Ag(+ or -Au) ore from veins hosted by skarn and recrystallized limestone. Horizontal metal zonation is well developed in the San Martin district. Cu and Ag correlate positively and the general pattern is Cu + Ag --> Cu + Zn --> Zn + Pb, with increasing distance from the intrusive contact. The contents of Fe, Cu, Zn, and Pb increase with depth within the ore zone. Au is farthest from the contact and occurs in veins within recrystallized limestone. Structural and stratigraphic controls were of major importance in localizing mineralization. Fractures in the Cuesta del Cura Formation associated with Laramide folding increased the permeability of the host rock; the metasomatic aureole, with accompanying sulfides, is most extensive in the most deformed portion of the limestone. Chert and shale units of the Cuesta del Cura served as local impermeable barriers to hydrothermal fluids; these units are mineralized only along fractures. The vein system represents a series of intrusion-related fractures that roughly parallel the intrusive contact and that served as major conduits for the ore-forming fluids. Formation of both the vein system and sulfide-hosting intercrystalline porosity in garnet skarn probably was aided by volume loss during metasomatism.Andraditie grandite garnet (average composition, Ca 3 Fe (sub 1.5) (super +3) Al (sub 0.5) Si 3 O 12 ) and hedenbergitie clinopyroxene (average, CaFe (sub 0.91) (super +2) Mg (sub 0.07) Mn (sub 0.02) Si 2 O 6 ) are the dominant prograde eale-silieate minerals; tremolite-aetinolite appears to be present as both a prograde and a retrograde phase. Other retrograde phases include wollastonite, vesuvianite, epidote, and ehlorite; fluorite and calcite are common, and minor quartz is also present. The metallic mineral assemblage is diverse and the paragenetie sequence can be divided into early, intermediate, and late stages. The sequence consists of early arsenopyrite, bornite, ehaleopyrite, pyrrhotite, and molybdenite; intermediate sphalerite, with intergrowths of ehaleopyrite, and galena; and late tetrahedrite-tennantite, pyrite, native silver, and stibnite. Supergene phases include mareasite, aeanthite, stromeyerite, and pyrargyrite.Deposition of grandite garnet probably was initiated by an increase in f (sub O 2 ) (and possibly decrease in. f (sub S 2 ) ) and took place at temperatures estimated in the range of 500 degrees to 550 degrees C. Garnet then became unstable relative to elinopyroxene and later eale-silieate alteration products. Fluid inclusion evidence suggests initially highly saline fluids (at least 24 wt % KCl and 36 wt % NaCl) with temperatures of major sulfide deposition starting at about 425 degrees C and declining thereafter. Initial values for f (sub O 2 ) and f (sub S 2 ) were such that metals were able to be transported as chloride complexes and sulfur was carried mainly as SO 2 . Sulfide precipitation was probably eansed by a continuing decrease in temperature and an increase in pH brought about by dissolution of CaCO 3 . Local endoskarn formation and noneconomic mineralization of the intrusion preceded exoskarn formation. Values for X (sub CO 2 ) in the hydrothermal fluids were low (< or = 0.05) throughout cale-silieate metasomatism and sulfide deposition. Relative metal solubilities were the major control on metal zonation. The Cu-Ag association is a product of thermal collapse of the mineralizing system, resulting in low-temperature mineral assemblages coexisting with high-temperature assemblages near the intrusive contact.