Circulation blood volume can be estimated with the initial distribution volume of glucose (IDVG) method. In a prospective, observational study, we evaluated the effect of a fluid challenge on IDVG in individual patients. In 13 patients after cardiac surgery, IDVG was determined before and after the infusion of 7 mL/kg hydroxyethyl starch over 30 min. Eleven patients not requiring a fluid challenge served as control. IDVG was computed with a one-compartment exponential model, using plasma glucose concentrations at baseline and 3, 4, 5, 6, and 7 min after 5 g of glucose IV bolus. IDVG was repeated 3 times in the control group. IDVG did not change after a fluid challenge (85 +/- 14 versus 93 +/- 14 mL/kg, P = 0.08), and the power of the comparison was 0.25 for a P value of 0.05. The regression coefficient of the exponential fit of glucose versus time was 0.96 +/- 0.03 before, and 0.95 +/- 0.04 after starch infusion (not significant). In the control group, IDVG was 90 +/- 18 mL/kg, and the average individual coefficient of variation was 0.15 +/- 0.08. IDVG seems inadequate to assess individual response to fluid therapy. This limitation may be related to the weak reproducibility of IDVG. CONCLUSIONS The initial distribution volume of glucose (IDVG) is supposed to reflect effective circulating blood volume. We assessed the sensitivity of this method to a fluid challenge in patients receiving a fluid challenge, and the reproducibility in otherwise stable patients. IDVG was not sensitive to a fluid challenge, and reproducibility was poor. Thus, clinical usefulness of IDVG seems limited.