OBJECTIVES: The study aimed to determine whether an inherent difference in vascular resistance between the liver and the ductus venosus influences the distribution of umbilical blood to the two sections of the circulation. STUDY DESIGN: Flow was measured in vitro in six fetal sheep liver preparations, perfused at varying pressures with saline solution or with blood of hematocrit 25% to 26% or 41% to 42%, with an occluded or open ductus venosus, at a temperature of 19° C to 21° C with likely absence of local control mechanisms. RESULTS: There was a linear relationship between pressure and flow through the liver, except at very low pressures when the linearity was broken. A closing pressure of 1 to 4 mm Hg was noted, lowest with saline solution and highest with hematocrit 42%. An open ductus venosus reduced overall resistance and brought down the closing pressure to <1 mm Hg. This also occurred during perfusion with high hematocrit. Thus at low pressures a proportionally greater fraction of umbilical flow was directed through the ductus venosus. CONCLUSIONS: The different physical properties of the portal vasculature and of the ductus venosus, pressure and viscosity, have a substantial impact on distribution of umbilical flow to the two parts of the circulation. Reduced umbilical venous pressure and increased hematocrit favor an increased fraction of flow through the ductus venosus.(Am J Obstet Gynecol 1997;177:90)