Kerrick and Caldeira (1993, 1994a) concluded that metamorphic CO 2 degassing in collisional orogens, and especially the Himalayan orogenic belt, could have been an important factor in enhancing paleoatmospheric CO 2 levels and contributing to early Cenozoic global greenhouse warming [Kerrick, D.M., Caldeira, K., 1993. Paleoatmospheric consequences of CO 2 released during early Cenozoic regional metamorphism in the Tethyan orogen. In: Touret, J.L.R., Thompson, A.B. (Guest-Eds.), Fluid–Rock Interaction in the Deeper Continental Lithosphere. Chem. Geol. 108, 201–230.] [Kerrick, D.M., Caldeira, K., 1994a. Metamorphic CO 2 degassing and early Cenozoic paleoclimate. GSA (Geol. Soc. Am.) Today 4, 57–65.]. However, our revised CO 2 mass loss computations for regional metamorphism in the Himalaya–Karakoram belt incorporating recent geochronologic data and revised estimates of the proportion of carbonate source rocks indicate that metamorphic CO 2 degassing from this orogen cannot explain Early Eocene warmth. Widespread pluton-induced hydrothermal flow occurred during the Eocene in the Cordilleran belt of western North America. Synmetamorphic intrusions, which are common in metamorphic belts, may cause significant regional fluid flow. To obtain a representative CO 2 flux from such environments, we computed a CO 2 flux of 1.5×10 12 mol km −2 Ma −1 from petrologic and geochemical studies of the Paleozoic plutonic–metamorphic belt in New England (northeastern United States). For the 2×10 6 km 2 area of Eocene metamorphism in the North American Cordillera, the CO 2 fluxes derived from the New England metamorphic belt yield an area-integrated flux of ∼3×10 18 mol Ma −1. If a significant fraction of this CO 2 entered the atmosphere, this degassing flux would alone account for Eocene greenhouse global warming. For the Ominica belt within the Cordilleran orogen, a volumetric estimate of the mass of carbonate veins indicates that the consumption of CO 2 by precipitation of carbonate veins may not significantly decrease the amount of CO 2 in fluids that convect to near-surface crustal levels. Compared to other Eocene metamorphic belts, the widespread hydrothermal activity in the North American Cordillera may have been the largest, and most climatically significant, source of metamorphic CO 2 to the Eocene atmosphere. CO 2 degassing by active metamorphism is most significant in extensional regimes of high heat flow. Extensional tectonism and hydrothermal activity in metamorphic belts may have substantially contributed to atmospheric CO 2 content throughout the Phanerozoic. Examples include the Mesozoic circum-Pacific metamorphic belt, and Oligocene–Miocene regional metamorphism in the Himalayan orogen.