Abstract An in vivo study was undertaken during hypothermic (28°C) cardiopulmonary bypass to compare oxygenator exhaust capnography as a means of estimating arterial carbon dioxide tension (P aCO 2) with bench blood gas analysis. A total of 123 pairs of measurements were made in 40 patients. Oxygenator exhaust capnography measurements systematically underestimated P aCO 2 measured by a bench blood gas analyzer. During the cooling and stable hypothermic phases of cardiopulmonary bypass, the relationship was reasonably accurate, but became far more variable during rewarming. Oxygenator exhaust capnography could be used as an inexpensive means of continuously monitoring P aCO 2 during the cooling and stable hypothermic phases of cardiopulmonary bypass but should not be used during rewarming.