This study was performed to assess the relation between Fick and indicator dilution measurements of cardiac output (CO) in a large number of subjects and to evaluate this relation in patients with a low CO, a high CO, and left-sided cardiac regurgitation. In 808 patients (428 men, 380 women, mean age 50 +/- 11), CO was measured by Fick and either thermodilution (right atrium to pulmonary artery)(n = 252) or indocyanine green dye ("dye")(pulmonary artery to systemic artery)(n = 556) within 10 minutes of each other. There was excellent agreement between Fick and both thermodilution and dye. The difference between Fick and indicator dilution measurements was 9 +/- 9%; it was 10% or less in 67% and 20% or less in 91% of patients. The disparity between Fick and indicator dilution measurements was increased in patients with a low CO (less than 2 liters/min/m2)(n = 152) (difference 14 +/- 11%, p less than 0.001) and those with aortic or mitral regurgitation (n = 83) (difference 13 +/- 11%, p less than 0.001). In these groups, the disparity between Fick and thermodilution measurements was not exaggerated, but the disparity between Fick and dye measurements was greater. Thus, although there is excellent agreement between Fick and both thermodilution and dye measurements of CO, thermodilution is preferable to dye in patients with a low CO and those with aortic or mitral regurgitation.