Calculations (with a 10-room and a simpler 3-room simulation of the large house fire) of temperature and smoke levels in several rooms of a structural fire are possible with the CFAST computer code. The accuracy and applicability of the results is greatly enhanced though the comparison of the calculations with experimental data. Experimental work thereby assists in understanding fire behavior in structural fires. Temperature measurements at different locations during a house fire provide necessary data for the development of mathematical models, which attempt to simulate the fire on a computer. In this paper, a large 170 square meter single-level house was subject to a complete experimental burn, with temperature measurements and fire observations during the entire burn, and subsequent modeling via a detailed 10-room simulation and a simpler 3-room simulation. The CFAST (Consolidated Model of Fire Growth and Smoke Transport) computer code is used to calculate temperatures and smoke levels in the various rooms of the house during the burn (with 10 different rooms). Four fire scenarios are considered in the simulation, with increasing realism regarding the actual fire specification. A simpler calculation (with 3 different rooms) has also done to see if the similar results would be shown with the 10-room simulation. It was found that results for smoke temperature and smoke layer heights were very similar, leading to the conclusion that a 3-room simulation of a 10-room building gives adequate modeling capability of the real structural fire. Computation results give the expected trends (deduced from local point temperature measurements) of initial temperature surge and decay, peak and leveling off temperatures, especially with respect to the northwest bedroom with a closed door. The effect of whether a door of a room would have been open was investigated computationally, with results illustrating far more dangerous smoke temperature and smoke level in the room when its door is open.