Composite stamping is a two steps process that includes an infrared heating oven in order to melt the composite sheets before forming. This study deals with the numerical simulation of the heating step of the process. The numerical model has been validated using three woven glass and carbon / PA6.6 composites provided by Solvay Rhodia. This type of simulation consists in solving the heat equation with a radiative flux that characterizes the interaction of the material with the IR heating. The model thus considers the IR properties of the material (emission and reflexion). Considering a homogeneous composite, the optical and thermal properties of sheets have been first measured. The material’s emissivity has been measured using a FTIR spectrometer from the reflective and transmitive spectra, by using the Kirchoff law and considering a Lambertian material. Three complementary measurement techniques were used to determine the thermal properties of the composites. Differential Scanning Calorimetry (DSC) measurements have been performed to identify the heat capacity of the composites. On another hand, a hot disc system (measurements performed at the LTN, France) has been used in transient conditions to determine the heat capacity and the thermal conductivity of the composites is all three directions. Finally, the in-plane thermal matrix of conductivity has also been measured by thermography by using an inverse method. The simulation of composites heating has been performed with Comsol MultiphysicsTM and the simulation procedure was validated by comparison with experimental results. The simulated IR oven is composed of 9 IR emittors provided by Toshiba Lighting Company that emit mainly in short IR wavelength (0.75-2µm). The emission properties of the tungsten filament were implemented in order to simulate the IR heating. Free convective heat transfer was also taken into account in the oven. In order to validate the model, an experimental set-up was instrumented with a calibrated IR pyrometer that measured the back side of the heated composite sheets. The experimental results confirm a low thermal gradient through composite thickness, in particular for carbon-reinforced composite. This result is consistent with the low Biot number of the composites. Moreover, experimental and simulated temperatures are in good agreement with an error lower than 15% in the entire heating stage from room to melt temperature.