The Steam-Assisted Gravity Drainage (SAGD) process has been successfully implemented to produce ultra-viscous bitumen from the Athabasca oil sands in the province of Alberta. In the Hangingstone area, 15 pairs of SAGD wells had been drilled in the reservoir by 2006, each a maximum of 30 m in thickness and approximately 300 m in depth. The production reached an average of 8,000 BOPD in recent years. The reservoir is geologically characterized as a stacked incised valley with fills in fluvial to upper-estuarine channels. Thin mudstone layers and abrupt changes in facies caused by sedimentary deposits present complexities and difficulties for SAGD implementation. A 3D seismic survey was conducted in 2002 to obtain a clear view of geology that was fully utilized for planning additional wells. In order to evaluate SAGD efficiency and performance, a time-lapse 3D seismic survey was carried out in 2006. In this paper, P-wave velocity (Vp) maps, transformed from the seismic travel-time maps, were interpreted with a new methodology for evaluating the areal extent of the steam chamber zone created by the SAGD process. In the previous experimental study of seismic velocity measurements with oil sands cores, Vp was found to steeply drop with an increase in temperature and to gently decrease with an increase in pore-pressure. Based on the experimental results, a petrophysical model was formulated to express Vp as a function of temperature, pressure and water saturation. The high-pressure and high-temperature zone of the SAGD process should generate differences between the first (2002) and second (2006) Vp maps from which we can estimate the area of the reduced bitumen viscosity with a temperature increase. As effects of pressure are probably more areally extensive than effects of temperature, these two effects on the Vp maps need to be segregated. As a new method, a scaling factor for the Vp reduction was first estimated to adjust the laboratory-scale and field-scale. We then calculated a distribution of Vp reduction corresponding to steam chamber conditions in order to decouple composite effects of temperature and pressure, based on the petrophysical model. Distinguishing the high-temperature and high pore-pressure zone from the low-temperature and high pore-pressure zone, we could determine a steam chamber distribution. The bitumen volume in the steam chamber zone was estimated and compared with the actual production. The methodology, interpretation procedures and results obtained are presented in detail.