Multiple processes support the significant efflux of carbon dioxide (CO2) from rivers and streams. Attribution of CO2 oversaturation will lead to better quantification of the freshwater carbon cycle and provide insights into the net cycling of nutrients and pollutants. CO2 production is closely related to O2 consumption because of the metabolic linkage of these gases. However, this relationship can be weakened due to dissolved inorganic carbon inputs from groundwater, carbonate buffering, calcification, and anaerobic metabolism. CO2 and O2 concentrations and other water quality parameters were analyzed in two data sets: a synoptic field study and nationwide water quality monitoring data. CO2 and O2 concentrations were strongly negatively correlated in both data sets (ρ = −0.67 and ρ = −0.63, respectively), although the correlations were weaker in high-alkalinity environments. In nearly all samples, the molar oversaturation of CO2 was a larger magnitude than molar O2 undersaturation. We used a dynamically coupled O2CO2 model to show that lags in CO2 air-water equilibration are a likely cause of this phenomenon. Lags in CO2 equilibration also impart landscape-scale differences in the behavior of CO2 between high- and low-alkalinity watersheds. Although the concept of carbonate buffering and how it creates lags in CO2 equilibration with the atmosphere is well understood, it has not been sufficiently integrated into our understanding of CO2 dynamics in freshwaters. We argue that the consideration of carbonate equilibria and its effects on CO2 dynamics are primary steps in understanding the sources and magnitude of CO2 oversaturation in rivers and streams.