The basic hypothesis of this study is that the intracranial aneurysm may enlarge and rupture due to dynamic instabilities of the blood flow and pressure inside the aneurysm. The specific question we attempted to answer is: which parameter(s) of aneurysmal geometry can serve as a reliable predictor(s) for aneurysmal rupture? We consider an idealized cylindrical aneurysm of the human common carotid artery and develop a mathematical model of blood flow through a normal artery and aneurysm connected in series. The mathematical model is nonlinear. It comprises nonlinear rheological properties of the normal artery and aneurysmal materials, and the inertial and resistance properties of the blood flow. The model equations were solved numerically and analyzed by methods of nonlinear dynamics. The critical aneurysmal diameter (CAD) is defined as a boundary point between the stable and unstable states of the model equations. The results confirm that a limit point of flow stability can occur only for a certain difference between aneurysmal and artery radii which are pre-disposed from a difference in their material properties. It was shown that CAD is dependent on both aneurysmal length and age of patient. Finally, the results suggest that the ratio between aneurysmal and normal artery diameters is a more reliable predictor of the aneurysmal rupture than the diameter alone. We conclude that an aneurysm diameter twice that of the normal artery could be dangerous.