Indocyanine green (ICG) has been demonstrated to be a reliable indicator for the estimation of plasma volume in man. Since this dye is bound to plasma protein when administered intravenously, overestimation of the plasma volume would occur during the presence of protein leak from plasma. However, to our knowledge, up to now no report has described whether or not plasma volume determined by the indocyanine green dilution method (PVICG) is clinically significantly overestimated under such conditions. We report two burned patients who temporally had an unusual large PVICG even though other clinical variables did not support the presence of hypervolemia. From the case described it can be concluded that a possible overestimation of PVICG occurs when the vascular permeability for proteins is increased.