Current models for predicting thermal sensation and thermal comfort as well as the solutions to improve the state of thermal well-being of the occupants of a building are insufficient. They do not sufficiently take into account interindividual differences of thermosensory judgment. However, these differences, often associated with thermal sensitivity of each person, exist but remain unexplained physiologically. This work, divided into two stages, is intended to identify the potential physiological causes of interindividual differences of thermal feeling through multiparametric experiments based on skin measurements. All measurements were performed after 30 minutes of acclimatization in controlled environment. The exploratory phase allowed to analyze both neurosensory activity, thermo-vascular properties and properties of the skin hydrolipidic film of two groups with different cold sensitivities (depending on their declared thermal sensation). For example, experiments have shown that it was more appropriate to analyse thermal and hydric skin properties (related to thermoregulation mechanisms) rather than neurosensory activity of volunteers to characterize interindividual differences of thermosensory judgment. They have also highlighted the need to control the non-thermal factors of environments and rigorously select subjects. The second step focused on the analysis of thermo-vascular properties and properties of the hydrolipidic film of two groups of different cold sensitivity. Thirteen women have faced in 6 environments of moderate temperatures between 17 ° C and 30 ° C (with 2 warm transitions and 2 cold transitions). Groups were built according to their degree of cold sensitivity. Differences in skin parameters have been found between the two groups. The most significant result is that cold-sensitive individuals have a more intense microcirculatory activity on cheeks with a stronger vasoconstriction in cold environments and a stronger vasodilation in warm environement than the non cold-sensitive group (p = 0. 002 according to ANCOVA test for groups effect). In addition, it has been shown that the multi-parametric approach (introduction of non-thermal parameters as predictors) as well as taking into account individual thermal sensitivities improve the prediction of thermal comfort especially for the cold-sensitive group (+ 6.4%).
