Adiabatic surface temperature is an efficient way of expressing thermal exposure. It can be used for bridging the gap between fire models and temperature models, as well as between fire testing and temperature models. In this study, a full-scale compartment fire experiment with wood crib fuel was carried out in a concrete building. Temperatures were measured with plate thermometers and ordinary thermocouples. Five plate thermometers and five thermocouples with a diameter of 0.25 mm were installed at different positions. These two different temperature devices recorded different temperatures, especially near the floor surface. The adiabatic surface temperature was derived by a heat balance analysis from the plate thermometer measurements. The thermal inertia of the plate thermometer was taken into account to correct the measured results. In addition, the fire experiment scenario was also simulated with fire dynamics simulator. The fire source was specified as a given heat release rate, which was calculated from the measured mass loss rate of the wood fuel. The adiabatic surface temperatures at these measuring positions were simulated by the fire dynamics simulator model and compared with the experimental adiabatic surface temperatures. The comparative results showed that fire dynamics simulator predicted the adiabatic surface temperature accurately during the steady-state period.