OBJECTIVE: Success in clinical trials of intracranial atherosclerosis (ICAS) is related to accurate identification of high-risk patients. Noninvasive computational fluid dynamics (CFD) of stenotic lesions may enhance therapeutic decision-making. We determined whether physiologic parameters change downstream cerebral hemodynamics in patients with ICAS. METHODS: Consecutive ICAS patients who underwent both CT angiography (CTA) and digital subtraction angiography were enrolled. CFD models were made using CTA source images. Inlet boundary conditions were defined using three ranges of systolic blood pressure (BP) (109.2, 158, and 225 mmHg) and hematocrit (27.3, 40.2, and 48.8). Ratios of pressure, shear strain rates (SSR), and flow velocity across the lesion were calculated using CFD simulations. A linear mixed model was used for the statistical analysis of repeated simulations. RESULTS: Among the 56 patients, 32 had moderate stenosis (50-69%) and 24 had severe stenosis (70-99%). A linear mixed model revealed that the ratio of pressure was predicted by systolic BP and stenosis group interaction (P =.036). These pressure decreases were associated with systolic BP (P <.001) and stenosis group (P <.001), but not with hematocrit (P =.337). Post-hoc analysis revealed that pressure decreases were more profound in the severe stenosis than the moderate stenosis group when comparing high and low systolic BP (P =.0108). Ratios of SSR and velocity were only associated in the stenosis group. CONCLUSIONS: Our study showed that systolic BP along with the degree of stenosis was associated with pressure decreases across stenotic lesions. Physiologic conditions may superimpose further changes in post-stenotic or downstream blood flow.