The purpose of this study is to measure hemodynamic changes in prefrontal cortex caused by acute cold in the right hand using fNIRS to calibrate the amount of pain caused by acute cold. fNIRS is a noninvasive functional brain imaging method which measures changes in concentration of HbO and HbR based on fNIRS principles. In this study, fNIRS signals of 21 adults (right-handed with no reported disease) were recorded in three stages. The fNIRS instrument used consisted of near and far channels on both sides of the forehead. Near channels measured hemodynamic changes in the superficial tissue and the far channels recorded changes in the cerebral cortex. By analyzing the data, two features were found to be effective in quantifying of the pain. The first feature was THb gradient during stimulation. Statistical analysis showed that THb exhibited significant changes in all channels during the first stage of CPT compared to the baseline. The second feature is the difference between maximum and minimum changes in THb concentration (ΔTHb); a significant relationship was observed between pain stimulus and ΔTHb in all channels. These results showed that fNIRS can be used as an effective tool to measure and study hemodynamic response caused by pain. Signals processing in the proposed algorithm includes: signal filtering, optical signal conversion to oxy and deoxyhemoglobin changes, elimination and normalization of signal trend. By extracting proper features from oxy, deoxy and THb by SVM and KNN classifiers, pain was classified in three different levels. The results indicate that Class 1 stimulus can be differentiated from Class 2 and Class 3 and also from their combination.