The current study was designed to evaluate a method for continuous measurement of cardiac output. The system consists of a modified pulmonary artery catheter that uses the thermodilution principle for determination of cardiac output. The evaluation was performed in vitro and in vivo. In-vitro evaluation was performed using a simple flow bench model (flow 2-9 L/min). Both continuous and bolus thermodilution methods were compared. Both methods showed good correlation with the pump flow calibrated using a volumetric tank and timer (correlation coefficient (r) for bolus thermodilution = 0.92, r for continuous thermodilution = 0.90). In-vivo evaluation was performed in six bovine experiments. Data from a total of 87 pairs of bolus versus continuous measurements were obtained. The cardiac output ranged from 1.9 to 8.9 L/min. The absolute measurement bias was not significant (mean: -0.07 L/min; 95% confidence limits: -0.87 and 0.73 L/min). The squared correlation coefficient from linear regression was 0.92. The results from this study suggest that the new continuous thermodilution measurement system for cardiac output provides accurate data in vitro and in vivo. Continuous monitoring of cardiac output adds a new dimension for evaluation of the patient's hemodynamic profile. Furthermore, significant volume load due to bolus thermodilution measurements can be avoided.