In general, cerebral blood flow accounts for 10–15% of cardiac output (CO), of which about 75% was delivered through the carotid arteries. Hence, if carotid blood flow (CBF) is constantly proportional to CO with high reproducibility and reliability, it would be of great value to measure CBF alternatively to CO. The aim of our study was to measure CBF in different cardiac cycles by ultrasound: Systolic carotid blood flow (SCF), Diastolic carotid blood flow (DCF) and Total (systolic and diastolic) carotid blood flow (TCF). And to investigate the direct correlation between CBF and CO. Whether the above correlation persisted when CO < 3.5/min. Patients aged 65 to 80 years old, undergoing elective cardiac surgery were included in this study. CBF and CO were measured by ultrasound and TEE respectively at 5 min, 10 min, and 15 min after induction of anesthesia. For all patients, the correlation coefficients between SCF and CO, TCF and CO were 0.41, 0.32 respectively, which were statistically significant, but not between DCF and CO. When CO < 3.5L/min, there was no significant correlation between either SCF, TCF or DCF and CO. It is suggested that systolic carotid blood flow should be used as a better index to replace CO. However, when the patient's heart function is poor, it is not recommended to use, and the method of direct measurement of CO is essential.