Introduction: Chronic liver disease is an important cause of morbidity and mortality. Almost 20% of patients have cirrhosis at presentation [1]. Previous studies have shown that alteration of hepatic perfusion parameters measured with CT or ultrasound in patients with liver cirrhosis correlate with the severity of chronic liver disease and the therapeutic outcome [2,3]. Recent studies assess liver cirrhosis by using color doppler velocity profile and contrast-enhanced sonography [4]. Early evaluation of the portal venous system with magnetic resonance velocity mapping 2D phase contrast MR angiography was applied [5]. Recent studies measuring degree of cirrhosis and portal hypertension with MR imaging and doppler US apply dynamic contrast material enhanced MR imaging [6]. The aim of this study was to provide improved diagnostic information by visualizing and quantifying comprehensive 3D vascular hemodynamics in the portal venous system using flow-sensitive 4-dimensional MRI at 3T. Methods: For the assessment of time-resolved 3D blood flow in the portal venous system, flow-sensitive 4D MRI was employed using a 3T MR system (TRIO, Siemens, Germany) and 3-dimensional MR velocity mapping in a group of 18 volunteers (age=28.6+/-3.1): venc =50cm/s, spatial res. = 1.6 x 2.1 x 2.4mm 3 , axial oblique 3D volume, 36 slices/slab, α = 7°, TE =3.0ms, TR = 44.8ms, temporal res. = 45ms. Respiration and wall motion artefacts were minimized by ECG and respiratory gating applying a navigator at the spleen-lung interface. The flow in the portal venous system was evaluated using 3D flow visualization (EnSight, CEI, Apex, USA) [7, 8]. Analysis was based on time-resolved 3D streamlines and particle traces originating from 6 emitter planes virtually positioned in the portal venous system (figure 1). All streamline and particle trace images were qualitatively graded according to the following categories: Visualisation of the vessels (2 = fully visible, 1 = partly visible, 0 = not visible), leakage into adjacent vessel system, maximum flow distribution, existence of vortices or retrograde flow, and type of inflow into the portal vein confluens (figure 1). Additional quantitative flow analysis using a home built tool (Matlab, TRhge Mathworks, USA) included interactive positioning of an analysis plane in the portal vein (see figure 1), vessel lumen segmentation, and flow quantification. For all volunteers flow waveforms over the cardiac cycle as well as peak velocity, minimal velocity and mean velocity were calculated (figure 2). Data in 15 volunteers were compared to the reference standard Doppler US (Hitachi EUB 7500 HV, Hitachi Medical Systems, Europe).