The non-invasive cardiac output (NICO) monitor is a new device in order to measure cardiac output (CO). A rebreathing circuit is built in the NICO monitor and CO is calculated using the Fick CO₂ equation. We compared this technique with the standard thermodilution (TDCO) technique in patients with thoracic and abdominal surgery. Thirty-two paired data were obtained in 17 patients. Correlation between the two methods in patients with controlled mechanical ventilation (CMV) was fair, with a correlation coefficient of 0.85. However, the correlation coefficient of the two methods was 0.60 in spontaneous breathing patients. Bland-Altman analysis showed a bias of 0.24±0.68 (mean±2SD) in CMV patients and 1.44±1.28 in spontaneous breathing patients. The NICO value was inversely proportional to an end-tidal CO₂ difference (ΔETCO₂) between pre-rebreathing and post-rebreathing. The large bias in spontaneously breathing patients might be due to a small ΔETCO₂ in spontaneously breathing patients. The NICO monitor has a tendency to measure higher CO values in spontaneously breathing patients.