Blood or plasma glucose concentration can be measured accurately and rapidly. However, after a glucose challenge metabolism may modify glucose kinetics, so that glucose has not been used as an indicator for dilution volumetry. To test the hypothesis that the initial distribution volume of glucose (IDVG) reflects cardiac output rather than glucose metabolism in the critically ill, the relationship between IDVG and thermodilution cardiac output was evaluated at 27 points in 13 non-surgical, critically ill patients without congestive heart failure. The IDVG was calculated from incremental plasma glucose concentrations using a one compartment model. Correlations were obtained between the IDVG and cardiac output (r = 0.89, n = 27, P < 0.001), and between the incremental plasma glucose concentrations three minutes after the injection and the IDVG (r = 0.94, n = 27, P < 0.001). No difference was found between the IDVG with or without continuous insulin infusions. The results indicate that the IDVG reflects cardiac output rather than glucose metabolism in patients without congestive heart failure.