We studied respirophasic variation in carotid artery blood flow peak velocity (ΔVpeak-CA) measured by pulsed wave Doppler ultrasound as a predictor of fluid responsiveness in mechanically ventilated patients with coronary artery disease.Forty patients undergoing elective coronary artery bypass surgery were enrolled. Subjects were classified as responders if stroke volume index (SVI) increased ≥15% after volume expansion (6 ml kg(-1)). The ΔVpeak-CA was calculated as the difference between the maximum and minimum values of peak velocity over a single respiratory cycle, divided by the average. Central venous pressure, pulmonary artery occlusion pressure, pulse pressure variation (PPV), and ΔVpeak-CA were recorded before and after volume expansion.PPV and ΔVpeak-CA correlated significantly with an increase in SVI after volume expansion. Area under the receiver-operator characteristic curve (AUROC) of PPV and ΔVpeak-CA were 0.75 [95% confidence interval (CI) 0.59-0.90] and 0.85 (95% CI 0.72-0.97). The optimal cut-off values for fluid responsiveness of PPV and ΔVpeak-CA were 13% (sensitivity and specificity of 0.74 and 0.71) and 11% (sensitivity and specificity of 0.85 and 0.82), respectively. In a subgroup analysis of 17 subjects having pulse pressure hypertension (≥ 60 mm Hg), PPV failed to predict fluid responsiveness (AUROC 0.70, P=0.163), whereas the predictability of ΔVpeak-CA remained unchanged (AUROC 0.90, P=0.006).Doppler assessment of respirophasic ΔVpeak-CA seems to be a highly feasible and reliable method to predict fluid responsiveness in mechanically ventilated patients undergoing coronary revascularization.NCT 01836081.