Summary Wellbore integrity is essential to ensuring long-term isolation of buoyant supercritical carbon dioxide (CO2) during geologic sequestration of CO2. In this paper, we summarize recent progress in numerical simulations of cement/brine/CO2 interactions with respect to migration of CO2 outside of casing. Using typical values for the hydrologic properties of cement, caprock (shale), and reservoir materials, we show that the capillary properties of good-quality cement will prevent flow of CO2 into and through cement. Rather, CO2, if present, is likely to be confined to the casing/cement or cement/formation interface. CO2 does react with the cement by diffusion from the interface into the cement, in which case it produces distinct carbonation fronts within the cement. This is consistent with observations of cement performance at the CO2-enhanced-oil-recovery Scurry Area Canyon Reef Operators Committee (SACROC) unit in west Texas (Carey et al. 2007). For poor-quality cement, flow through cement may occur and would produce a pattern of uniform carbonation without reaction fronts. We also consider an alternative explanation for cement carbonation reactions as caused by CO2 derived from caprock. We show that carbonation reactions in cement are limited to surficial reactions when CO2 pressure is low (< 10 bar), as might be expected in many caprock environments. For the case of caprock overlying natural CO2 reservoirs for millions of years, we consider the Scherer and Huet (2009) hypothesis of diffusive steady state between CO2 in the reservoir and in the caprock. We find that, in this case, the aqueous CO2 concentration would differ little from that in the reservoir and would be expected to produce carbonation reaction fronts in cements that are relatively uniform as a function of depth.