BACKGROUND: The incidence of obesity and hypertension, with all their associated complications, is on the rise globally. Non-invasive blood pressure (NIBP) is often difficult to measure in obese individuals owing to the increased circumference and conical shape of the upper arm. The forearm and ankle may serve as convenient alternatives to measure NIBP. The aim of this study was to identify the most accurate site for NIBP monitoring in obese patients. OBJECTIVES: To statistically determine an agreement between different NIBP sites (upper arm, forearm and ankle) and our control blood pressure (BP). A secondary objective was to investigate agreements between different NIBP sites, anthropometric variables (body mass index (BMI)/mid-upper arm circumference/conicity index) and the control, with the goal of deriving a correction formula for BP. METHODS: A prospective cross-sectional study was conducted at a provincial tertiary hospital in Gauteng Province, South Africa. Fifty participants aged 20 - 60 years, with BMI 35 - 40 kg/m2, were recruited. Using appropriately sized cuffs, NIBP measurements were obtained from the left upper arm, forearm and ankle. Simultaneous continuous BP measurements were recorded using the VitalStream device on the contralateral hand. Bland-Altman plots and regression analyses were employed to evaluate agreement and derive correction formulas for each measurement site. RESULTS: Bland-Altman analysis revealed significant biases across sites, with ankle systolic BP showing the greatest deviation (mean bias +14.44 mmHg, 95% confidence interval 8.26 - 20.62). Regression analyses identified significant agreements for mean arterial pressure at the upper arm, enabling a correction formula with high reliability (p<0.001). Forearm and ankle measurements demonstrated wider limits of agreement and were prone to overestimation, especially in systolic and diastolic pressures. CONCLUSION: Upper arm NIBP measurements were the most accurate in the obese population. Alternative sites, such as the forearm and ankle, demonstrated inconsistent reliability, and require careful interpretation. Correction formulas can enhance the accuracy of NIBP readings but may be cumbersome for routine clinical use. Future studies should focus on refining measurement protocols and evaluating the efficacy of conically shaped cuffs for improved accuracy.