There is currently no practical method for determining cardiopulmonary resuscitation (CPR) efficacy in the field. We investigated the relationship between the volume of carbon dioxide (CO2) excreted in the airway (CO2EX) when tidal volume and respiratory rate are controlled, and cardiac output (CO), an indicator of CPR efficacy, to determine the potential of CO2EX as a practical noninvasive field monitor of CPR efficacy. Thirteen mongrel dogs were anesthetized, instrumented and ventilated 13 times/min at a fixed tidal volume with an infrared airway CO2 sensor. CO2EX = (PCO2/bar. press) x (tidal vol) x (breaths/min), and expressed in ml/min per kg. Sequences of control, CPR with 3-4 different compression forces, and recovery measurements were recorded 10-15 times/animal. CO2EX and CO fell simultaneously with ventricular fibrillation. CPR immediately increased CO2EX and CO. Both changed consistently and in the same direction as compression force. Return of spontaneous circulation immediately increased CO2EX and CO above controls, with a gradual return to control levels. CO2EX was always below 8 ml-CO2/min/kg during CPR and above this during spontaneous circulation. With alveolar ventilation controlled, CO2 movement is regulated by CO, CO distribution and CO2 stores shifts. Normally, CO accounts for 15% of CO2EX variability. In this study CO accounted for > or = 65% of CO2EX variability during CPR, indicating CO2EX changes were primarily due to CO changes. When ventilation is controlled, CO2EX during CPR reliably tracks changes in CO. Therefore, CO2EX may provide a practical noninvasive method of determining CPR efficacy as the CPR is being performed.