Background We demonstrate a new formula to predict mean arterial pressure (MAP) using corrections of the key factors associated with the inaccuracy of the standard formula: heart rate (HR) and pulse pressure (PP). Methods A total of 99 patients (50 men, 49 women; mean age 52.5 ± 10.3 years), who underwent elective coronary angiography, were enrolled in our study. The arterial pressure was measured in the aortic root. MAP was measured digitally by the area-under-the-pressure-time curve method. We evaluated the accuracy of four different formulas: the standard formula, the formula of Razminia et al ., the formula of Meaney et al . and the combined formula. Results PP coefficient deviation of the standard formula was negatively correlated with PP ( R = −0,561, P < 0.001), and positively correlated with HR ( R = 0,298, P = 0.003). Both R and R 2 values of the combined formula were higher than previous formulas. Accuracy parameters of the combined formula [root mean square error (RMSE): 1.801 mmHg, mean square residuals (MSR): 3.244 mmHg 2 , Akaike information criterion (AIC): 401.4] were superior to the standard formula (RMSE: 1.902 mmHg, MSR: 3.620 mmHg 2 , AIC: 412.3), the formula of Razminia et al . (RMSE: 2.022 mmHg, MSR: 4.089 mmHg 2 , AIC: 424.3) and the formula of Meaney et al . (RMSE: 2.137 mmHg, MSR: 4.568 mmHg 2 , AIC: 435.3). In the multivariate linear regression analysis, the combined formula was the only method that independently predicts the measured MAP (beta = 0,990, P < 0.001). Conclusion The combined formula is superior to previous formulas for accurately predicting MAP.