BACKGROUND: Internal carotid artery (ICA) blood flow may be an alternative target to measure cerebrovascular reactivity (CVR) and endothelium-dependent vasodilatory dysfunction. We investigated the feasibility of testing CVR in the ICA during routine clinical carotid investigation. METHODS: CVR was assessed by effects of breath-hold (BH) and rapid-breathing (RB) on ICA diameter, blood velocities (peak systolic velocity [PSV], end diastolic velocity [EDV], and mean velocity [MV]), volume flow, and relative to changes in end-tidal carbon dioxide (end-tidal CO2 [etCO2]; CVRPSV, CVREDV, CVRMV, and CVRFlow). Feasibility and validity were determined by data quality, agreement between challenges, and association with age. RESULTS: In 50 patients (68 ± 13 years old, 15 female), recordings during rest, BH, and RB were of good quality in 96%, 90%, and 88%, respectively. Age-related reductions in blood velocities were evident during rest (EDV: r 2 = 0.37, P < 0.0001), BH (EDV: r 2 = 0.42, P < 0.001), and RB (EDV, r 2 = 0.50, P < 0.0001). During BH, the absolute changes in blood velocities (EDV: r 2 = 0.15, P < 0.01), etCO2 (r 2 = 0.10, P = 0.03), CVREDV (r 2 = 0.13, P = 0.02), and CVRMV (r 2 = 0.11, P = 0.03) were smaller in older patients. In contrast, during RB, absolute changes in blood velocities and etCO2 were not age related, but changes in volume flow (r 2 = 0.12, P = 0.02), CVRMV (r 2 = 0.11, P = 0.03), and CVRFlow (r 2 = 0.16, P < 0.01) were greater in older patients. CONCLUSION: During routine carotid duplex assessment, assessing CVR in the ICA was feasible, practical, and associated with high quality recordings. Differential associations with age between hypercapnic and hypocapnic stimuli imply a greater tendency to vasoconstriction in older patients.