The perfusion of the liver is complex, especially at the microcirculatory level. The functional units of the liver are often schematized as hexagonal lobules (Fig. 1a), which receive blood from the peripheral portal triads (PT; including hepatic arterioles and portal venules). A part of the PT blood drains into the vascular septa (VS), which are the vascular beds delineating the lobule boundaries in between successive portal triads. Subsequently, the blood enters the interconnected network of tortuous sinusoids, where the metabolic exchange with neighboring hepatocytes takes place. Afterwards, blood drains radially into the central vein. Despite this well-known conceptual model, liver microcirculation is still not fully understood. Previously, the liver microhemodynamics have been modeled using simplified (2D) geometries and/or a porous media approach with an isotropic permeability. However, the validity of these assumptions has never been assessed. Therefore, the aim of this study was threefold. First, the (an)isotropic permeability behavior of an image-based 3D sinusoidal network was quantified using computational fluid dynamics (CFD). Secondly, the resulting permeability tensor was applied to build a 3D CFD porous medium model of a liver lobule. Thirdly, the role of VS in hepatic microperfusion was investigated by comparing a porous lobule model with and without VS.