BACKGROUND End-stage renal disease (ESRD) is complicated by high blood pressure (BP) and severe atherosclerosis, which may lead to different values of BP measured by invasive and noninvasive methods. However, the differences in BP values between these two methods and their possible mechanisms in ESRD have not been investigated. METHODS Forty-one ESRD patients on dialysis were enrolled in the present study. Invasive BP, noninvasive BP, and carotid hemodynamic variables were obtained through radial artery catheterization, oscillometric methods, and echoarteriography, respectively. RESULTS We found that noninvasive systolic BP (NISBP) was 30±20 mmHg (P<0.001) lower than invasive systolic BP (INSBP), whereas no significant difference was found in diastolic BP between these two methods. The difference between INSBP and NISBP was correlated with carotid peak systolic velocity (r=0.3550, P=0.0229). CONCLUSION NISBP could not accurately represent INSBP in ESRD patients on dialysis. The lack of dynamic pressure measurement during NISBP measurement may contribute to the difference between NISBP and INSBP.