Theoretically, systemic carbon dioxide (VCO2) production should be an alternative means to systemic oxygen uptake (VO2) for estimating the global efficacy of cardiopulmonary bypass (CPB). This study compared two methods of estimating VCO2: Fick's principle and oxygenator exhaust carbon dioxide (CO2) output. Both of these estimates were then compared with VO2. Fifty-one patients (39 male and 12 female) undergoing elective cardiac surgery requiring CPB were studied. Blood sampling was performed and measurements recorded during active cooling, environmental cooling/stable hypothermia and during rewarming. Blood samples were measured for CO2 tension from which content was estimated. VCO2 was calculated as the product of the arteriovenous difference in CO2 content and pump flow rate (Fick's principle), or the fresh gas flow rate and concentration of the oxygenator exhaust CO2 (output technique). Over all measurements, method comparison analysis revealed a large mean bias of 41 (95% confidence intervals (CI) 32-50) mL/min with very wide limits of agreement (-23, 105 mL/min). Regression analysis found that the bias was also proportional to the size of measurement (beta = 0.75 (95% CI 0.55, 0.95)). Although both methods of VCO2 correlated significantly with VO2 (p < 0.01), regression analysis found that the coefficients (beta) of both techniques had wide CI (Fick's principle: beta = 1.37 (95% CI 1.20, 1.54); output technique: beta = 0.58 (95%CI 0.44, 0.71)). In conclusion, both techniques of VCO2 cannot be used interchangeably, and both are imprecisely related to VO2 as estimated by Fick's principle.