The western North American Cordillera hosts a large number of gold-bearing quartz vein systems from the Mother Lode of southern California, through counterparts in British Columbia and southeastern Alaska, to the Klondike district in central Yukon. These vein systems are structurally controlled by major fault zones, which are often reactivated terrane-bounding sutures that formed in orogens built during accretion and subduction of terranes along the continental margin of North America. Mineralization ages span mid-Jurassic to early Tertiary and encompass much of the evolution ofthe Cordilleran orogen. Nitrogen contents and δ15N values of hydrothermal micas from veins are between 130 and 3,500 ppm and 1.7 to 5.5 per mil, respectively. These values are consistent with fluids derived from metamorphic dehydration reactions within the Phanerozoic accretion-subduction complexes, which have δ15N values of 1 to 6 per mil. The δ18O values of gold-bearing vein quartz from different locations in the Cordillera are between 14.6 and 22.2 per mil but are uniform for individual vein systems. The δD values of hydrothermal micas are between -110 and -60 per mil. Ore fluids have calculated δ18O values of 8 to 16 per mil and δD values of -65 to -10 per mil at an estimated temperature of 300δC; δD values of ore fluids do not show any latitudinal control. These results indicate a deep crustal source for the ore-forming fluids, most likely of metamorphic origin. Low δDH2O values of -120 to -130 per mil for a hydrous muscovite from the Sheba vein in the Klondike district reflect secondary exchange between recrystallizing mica and meteoric waters. Collectively, the N, H, and O isotope compositions of ore-related hydrothermal minerals indicate that the formation of these gold-bearing veins involved dilute, aqueous carbonic, and nitrogen-bearing fluids that were generated from metamorphic dehydration reactions at deep crustal levels. These data are not consistent with either mantle-derived fluids or granitoid-related magmatic fluids, nor do they support a model involving deeply circulated meteoric water.