Thermal bridges tend to increase overall buildings energy demand and might cause water condensation problems. They are thermally characterized by a linear transmission coefficient ψ or a point transmission coefficient χ. Today, most studies of thermal bridges are based on theoretical or numerical calculations. Standardized methods define default values and assumptions to make in simple or detailed simulations. The few existing in situ characterization methods of thermal bridges are based on steady-state assumptions. This makes them highly sensitive to weather conditions and often requires very long measurements. The present paper proposes a novel active method for the in situ characterization of a thermal bridge. It generalizes a measurement of a homogeneous wall thermal resistance. The indoor air is rapidly heated for a few hours (typically 6) and the wall thermal response is analyzed with an inverse technique based on a white-box model. Surface temperatures and heat fluxes are measured with contact sensors on a sound area and these quantities are then extrapolated to nearby thermal bridges using infrared thermography. The total heat transfer coefficient, required in the heat flux extrapolation process, is monitored with a specific device. The method is validated on a full-size loadbearing wall built inside a climate chamber. The mechanical supports, holding the internal insulation system implemented on the wall, generate several thermal bridges. The ψ-values estimated by the active method are less than 20% away from steady-state measurements taken as reference. Many configurations were tested with constant and varying external temperature and the method proved its robustness to these unsteady conditions.