An experimental set-up of triple-glazed supply-air window is developed in this research in order to characterize the window’s thermal performance without solar radiation. By simultaneously measuring the local temperatures and heat fluxes with thermocouples and heat fluxmeters, the heat fluxes determination requires no longer using the correlations of heat transfer coefficients around the window, which are often the source of high uncertainties. Results show that the use of fluxmeters brought a more accurate measure of heat transfers around and in the window. Thereafter, the heat transfer coefficients can be correctly estimated by empirical evidence. Uncertainty analysis is then presented to highlight the reliability of the experimental method. Afterwards, the obtained experimental data are compared with those of numerical model developed by using Fluent® software. A thorough comparison analysis is provided to explain which parameters play a role in deviating the results between the two methods, leading to conclude the validity of numerical model assumptions with respect to the real conditions of experimental set-up.