This paper investigates the correlation of human thermal sensation with multiple physiological information in stable indoor thermal environment. Fifteen young male subjects from East China participate in the experimental study. The thermal sensation votes (TSVs) are collected by subjective questionnaire, and six physiological information are measured simultaneously, i.e., wrist skin temperature, blood pressure, heart rate, heart rate variability (ECG), skin conductance and blood oxygen saturation. A total of nine different physiological markers including four ECG signals are recorded. The deviation analysis between TSV and PMV standard is conducted. A detailed correlation analysis of nine physiological markers with thermal sensation is performed. It is found that young males of East China prefer lower ambient temperature in summer than PMV’s prediction. Two environmental and seven physiological variables have obvious correlations with TSV. Among them, wrist skin temperature, heart rate and ECG (pNN50) are three physiological markers that have high correlations with TSV. On this basis, a data-driven thermal sensation predictive model based on multiple physiological information is designed. The accuracy is better than that of skin temperature based predictive model or PMV evaluation.