Objective: To evaluate potential changes in the ratio of whole-body/large-vessel hematocrit (f-cell ratio) during isovolemic hemodilution and to compare the volume effects of 2 different plasma exchange solutions (hydroxyethyl starch 200,000/0.62 6% and human albumin 5%). Design: Prospective, randomized, controlled trial. Setting: Operating theater in a university hospital. Patients: 24 gynecological patients scheduled for elective surgery. Interventions: Isovolemic hemodilution was performed using 2 different plasma exchange solutions. Plasma volume was determined using dye dilution technique before and after hemodilution. The volume of withdrawn blood was measured from the change in weight of the blood bags taking into account the specific gravity of blood. Results: The volume of administered plasma exchange solutions exceeded the amount of withdrawn blood by 80 ± 47 ml (p < 0.001). Plasma volume was 3,067 ± 327 ml before and 3,517 ± 458 ml after hemodilution. Using red cell volumes calculated from measured plasma volumes and peripheral hematocrit, a deficit of 249 ± 133 ml (p < 0.0001) in red cells after hemodilution appeared with the measured withdrawn red cell volumes taken into account. This finding can be explained by a change in the f-cell ratio during isovolemic hemodilution. The volume effect of the exchange solutions was 1.05 for hydroxyethyl starch and 0.95 for albumin. Conclusions: The results demonstrate that a change in the f-cell ratio occurs during isovolemic hemodilution. The estimation of red cell volume or plasma volume changes by using either the hematocrit or plasma or red cell volume determinations together with the hematocrit may lead to erroneous results.