The purpose of this study was to relate local blood flow conditions to the development of poststenotic dilatation (PSD) in the common carotid arteries of rabbits. We investigated the effects of the geometry of the stenoses on PSD formation to gain insight into the mechanisms of this phenomenon. With short stenoses, the maximum diameter of the PSD after 3 weeks increased from about 25% to 50% above the proximal diameter by increasing the severity of the stenoses from 50% to 60% diameter reduction. By contrast, increasing the length of the constricted region of 60% stenoses did not affect the PSD, and in all cases, the site of maximum dilatation occurred within 3 tube diameters from the stenoses. In vitro studies were conducted with the photochromic tracer technique. The most striking observation was that the transition to turbulence did not occur with a 55% short stenosis. By increasing the severity of this short stenosis to 70%, the transition to very localized turbulence was triggered =6 to 8 tube diameters from the stenosis during the early deceleration phase of the flow cycle. The transition point moved farther downstream by increasing the length of this moderate stenosis. This study demonstrated that PSD can occur under turbulence-free conditions, and even when turbulence was generated, the site of the PSD was remote from that of the localized turbulence zone. The wall shear stress pattern was determined for the long 60% stenosis. When compared with the values measured upstream from the stenosis, the wall shear stress within the throat was estimated to be at least 30 times larger, whereas downstream from the stenosis, it was roughly two times larger, together with strong cycle-to-cycle variations. We hypothesize that extreme magnitudes of the shear stress or shear stress variations in the throat of the stenosis or in the poststenotic region can give rise to PSD formation, probably by stimulating the release of agents that induce vessel wall remodeling when delivered to tissues around the poststenotic primary vortex. It is more likely that shear stresses in the stenosis throat than in the downstream region are important, since both shear stresses at the throat and the degree of PSD were sensitive to degree of stenosis, whereas downstream shear stresses were not PSD is not initiated by shear-induced release of endothelium-derived relaxing factor; experiments with an inhibitor of endothelium-derived relaxing factor synthesis, the argjnine analogue /VG-nitro-L-argiiiine-methyl ester, failed to influence the development of PSD.