Measurement of cardiac output (CO) is of importance in the diagnostic of critically ill patients. The invasive approach of thermodilution (TD) via pulmonary artery catheter is clinically widely used. A new non-invasive technique of inert gas rebreathing (IGR) shows a good correlation with TD measurements in spontaneously breathing individuals. For the first time, we investigated whether IGR can also be applied to sedated and mechanically ventilated subjects with a clinical point of care device. CO data from IGR were compared with TD in six healthy mongrel dogs. Data sampling was repeated under baseline conditions (rest) and under stress challenge by applying 10 μg/kg/min of dobutamine intravenously. Switching from mechanical ventilation to IGR, as well as the rebreathing procedures, were carried out manually. Cardiac output data from IGR and TD correlated with a coefficient of r=0.90 (95% confidence interval [0.81; 0.95]). The Bland-Altman analysis showed a bias of 0.46 l/min for the IGR CO measurements. Ninety-five percent of all differences fall in the interval [-1.03; 1.95], being the limit of the ± 1.96 standard deviation lines. IGR is a new approach for non-invasive cardiac output measurement in mechanically ventilated individuals, but requires further investigation for clinical use.