OBJECTIVE: To determine whether indocyanine green (ICG) and glucose dilutions can detect generalized capillary protein leakage in septic patients without requiring repeated measurements. DESIGN: Prospective, clinical study. SETTING: General intensive care unit. PATIENTS: Twelve consecutive patients who met the criteria of sepsis and 16 consecutive acute myocardial infarction (AMI) patients without any underlying pathology inducing generalized protein capillary leakage. INTERVENTIONS: Both ICG 25 mg and glucose 5 g were administered simultaneously, to calculate the plasma volume determined by the ICG dilution method (PV-ICG) and the initial distribution volume of glucose (IDVG), on day 1 of sepsis or on day 1 of hospitalization for the AMI patients. The relationship between these two volumes and the PV-ICG/IDVG ratio was evaluated in two patient groups. MEASUREMENTS AND MAIN RESULTS: Although the IDVG of the two patient groups was not statistically different, the PV-ICG in the septic patients was higher than that in the AMI patients (p < .01). Consequently, the PV-ICG/IDVG ratio in the septic patients was higher than that in the AMI patients (p < .01). Eight of the 12 septic patients had a PV-ICG/IDVG ratio of >0.45, which was not observed in any of the AMI patients. The PV-ICG/IDVG ratio in the septic patients correlated inversely with the total plasma protein concentration (r2 = .46, p < .025) and mean arterial pressure (r2 = .42, p < .05). CONCLUSIONS: Our results indicate that overestimation of the PV-ICG can occur in septic patients and, further, suggest that simultaneous measurement of the two distribution volumes would help predict generalized capillary protein leakage in septic patients without repeated measurement.