Polyurethane Foams are produced by mixing Polyol, Isocyanate and water. Two chemical reactions are then in competition: Isocyanate and water produce CO2 which governs the development of bubbles. The crosslinking reaction between isocyanate and polyol produces the solidification of the polyurethane skeleton. Obviously the quality of the foam (bubble size and density) will depend on the coupling between these two reactions. If the crosslinking reaction starts too early, the viscosity of the skeleton will increase and block the foam expansion. On the other hand, if the crosslinking reaction starts too late, the foam will collapse because the structure of the skeleton is not strong enough. As a consequence, the viscosity of the material is a key parameter to master the foaming process as well as the characteristics of the foamed product. It is very difficult to measure the viscosity of the (polyol-isocyanate-water) mixture because the time scale of CO2 production and crosslinking are very short and hardly compatible with the installation of the sample on a classical parallel plate rheometer. A Foamat system (Format Messtechnik GmbH) is used to get experimental data. It consists in a tubular mold equipped with thermocouple and pressure transducer. The mixture is introduced in the mold and the foaming dynamics (the height of the sample as a function of time) is measured simultaneously with the temperature and the pressure development. An analytical thermomechanical model of the foaming process inside the Foamat is proposed. It allows identifying the viscosity as a function of time during the foaming process. Nevertheless this analytical model is based on severe assumptions (temperature and foam density homogeneous in the mold at each time step during the foaming process). A numerical computation using REM3D and the identified rheology functions allows to justify the analytical developments.