Thermal comfort, influenced by thermal sensation is an important building performance indicator. In this study we discuss the use of a thermophysiological model in the built environment to assess thermal sensation. In the context of this work, the use of CFD to simulate the thermal environmental conditions around a human is analyzed. Experimental data from two independent studies, covering both genders, are used to validate three different, currently available, thermal sensation models: (1) the Predicted Mean Vote index (PMV), (2) the UC Berkeley thermal sensation model and (3) the EN-ISO 14505 standard. Use of such a model is required to link physiological responses to thermal sensation. In this study they have been evaluated for two different steady-state non-uniform thermal environments. The results confirm that the PMV is not capable of predicting whole body thermal sensation when local effects (local skin temperatures and thermal sensation) have a significant influence. The results furthermore indicate that the use of a thermophysiological model (ThermoSEM) in combination with the UC Berkeley model or EN-ISO 14505 standard seems to be promising regarding the prediction of thermal sensation of local body parts and overall thermal sensation under steady-state non-uniform environments. The advantage of using a thermophysiological model in combination with a thermal sensation model is that thermal comfort can be assessed on a more individualized level under complex, daily encountered, thermal environments where local effects play an important role. However, both thermal sensation models need more research before they can be used in daily building design practice.