Over the past few years, a metabolic rate meter has been introduced for easy measurement of oxygen consumption. However, its accuracy is unproved. In 40 patients (26 men, 14 women, ages 34 to 73 years), cardiac output was measured simultaneously by thermodilution and the Fick method using the metabolic rate meter to quantitate oxygen consumption. In comparison with thermodilution, the results using the Fick method were low (5.26 +/- 1.18 vs 4.14 +/- 0.99 liters/min, respectively, p less than 0.01). In 18 patients cardiac output also was measured by the Fick method using a Douglas bag to quantitate oxygen consumption. In these patients, oxygen consumption measured with the metabolic rate meter was lower than that obtained using the Douglas bag (168 +/- 25 vs 216 +/- 42 ml/min, respectively, p less than 0.01). With the Douglas bag, the Fick and thermodilution cardiac output measurements were similar (4.68 +/- 1.08 vs 4.87 +/- 0.86 liters/min, respectively, difference not significant), and they differed by less than or equal to 10% in 15 patients. In contrast, with the metabolic rate meter, the results of thermodilution were higher than those with the Fick method (4.84 +/- 0.95 vs 3.60 +/- 0.71 liters/min, respectively, p less than 0.01), and differed by less than or equal to 10% in only 1 patient (p less than 0.01). Thus, the values for oxygen consumption and cardiac output obtained with the metabolic rate meter are lower than actual values. This device is less accurate than the Douglas bag.