Fanger's heat balance model led to the formulation of the predicted mean vote (PMV) and predicted percentage dissatisfied (PPD) for rating thermal comfort in buildings. The authenticity of PMV/PPD model has been questioned by several studies, thus opening the debate for revisiting the main parameters in PMV thermal comfort model. In this experimental study, a correlation between biomarker �mean blood pressure (MAP)� and the �activity level�, was developed to improve the thermal prediction of PMV model. This study revealed a strong correlation between mean blood pressure and the activity level with a confidence level of 96. Field assessments of PMV model were conducted in air conditioned as well as naturally ventilated buildings to analyze the effect of mean blood pressure on the PMV model. In air-conditioned buildings, PMV model overestimated the thermal sensation up to 54 as compared to actual vote, whereas the overestimation of modified model (mPMV) was found to be 22 only. The PPD deviations of mPMV and PMV models were found to be 8 and 28 respectively. Statistical analysis on the collected data strengthened the significance of mPMV on PMV model. In naturally ventilated buildings, the correlation found to be insignificant due to uncontrolled variables. © 2016 Elsevier Ltd