Purpose : The most important factor discrediting the reliability of MRAs is the overestimation of the degree of stenosis in the internal carotid artery(ICA). The purpose of this study is to evaluate the secondary hemodynamics and the cause(s) for the overestimation of the degree of variable stenotic phantoms of the carotidartery using steady-state flow on MRAs. Materials and Methods : Using acrylic materials, normal and variable stenotic phantoms of the bifurcated carotid artery were constructed (40% and 65%). Flow patterns were evaluated with axial and coronal imaging of MRAs (2D-TOF and 3D-TOF) and DSAs of phantoms constructed from an automated closed-type circulatory system filled with 10% glucose solution. These findings were then compared with those obtained from CFD. Results : 3D-TOF axial MRA of asymmetrically 40 percent stenotic phantom revealed 40 percent stenosis identical to the stenotic region of phantoms with continued poststenotic signal loss, whereas 3D-TOFzsial MRA of symmetrically 65 percent stenotic phantom showed markedly decreased signal intensity at the poststenotic segment resembling occlusion. Source image of 2D-TOF coronal MRA showed redistribution (from the internal to external carotid artery side) of the central axis of inflow depending upon the degree of stenosis of the ICA ; this redistribution can be a cause of the decreased signal at the poststenotic segment, due to a reduced volume of flow through the stenotic segment. The general hemodynamics of the variable stenotic phantoms on MRA were identical to the hemodynamics on DSA and CFD. Conclusion : Although dephasing from turbulent flow and character of maximum intensity projection (MIP) were suggested as the main cause of the decreased poststenotic signal, our study indicated that a hemodynamically redistributed central axis of inflow and reduced flow volume through stenotic channel is one of the basic factors of the decreased signal intensity at the poststenotic segment on MRA.