Wood studs protected with gypsum plasterboards have been tested in the cone calorimeter to investigate the possibility of using a conical radiator to predict furnace behaviour for light timber frame walls. Other phenomena, such as the behaviour of gypsum boards, insulation materials and wood when exposed to high heat flux, have also been of interest during the experimental and theoretical work. It has been shown that it is not possible to use a constant heat flux in the cone calorimeter to achieve the similar charring rate and time to onset of char as those observed in full-scale experiments. The results from the cone calorimeter tests have been compared with predictions from a finite element method heat transfer software using thermal properties recommended in the literature. The model has thereafter been calibrated to give better predictions by modifying the thermal properties of gypsum. The calibrated model has been used to simulate furnace tests. The simulation results have later been compared with the measured full-scale data. It has been concluded that a computer model can be used to calibrate the thermal properties to agree better with the experimental temperature measurements in the wood. The simulated charring rate corresponded well with the charring rates that were measured in the furnace tests but the model overpredicted the time to onset of char.