Despite recent evidence suggesting that groundwater inputs of dissolved inorganic carbon (DIC) to rivers can contribute substantially to the fluvial evasion of carbon dioxide (CO2), groundwater is seldom integrated into fluvial carbon budgets. Also, unclear is the way equilibria between CO2 and ionic forms of carbonate will affect CO2 evasion from rivers. We conducted longitudinal river surveys of radon and carbon along two rivers of tropical Australia and developed a mass balance framework to assess the influence of groundwater‐derived inorganic carbon and carbonate buffering on CO2 evasion rates. The mean CO2 evasion flux totaled 8.5 and 2.3 g·C·m−2·day−1 for the two rivers, with considerable spatial variations that we attributed primarily to changes in groundwater inflow rates (minima and maxima per river reach 1.2–45.1 and 0.2–13.4 g·C·m−2·day−1). In the larger river system, inflowing groundwater delivered on average 6.7 g·C·m−2·day−1 as dissolved CO2—almost 10 times as much as the CO2 produced via river metabolism—and 21.6 g·C·m−2·day−1 as ionic forms. In both rivers, these groundwater‐derived inputs were a mixture of biogenic and geogenic carbon sources. Spatialized estimates of the carbonate buffering flux revealed that in reaches where CO2 evasion was particularly high, the carbonate system was able to maintain high CO2 concentrations by adjustment of carbonate equilibria. This process was likely triggered by high groundwater inflow rates. Our findings suggest that both groundwater inputs and carbonate equilibria need to be accounted for in fluvial carbon budgets, particularly in high‐alkalinity rivers.