Four thermal processes (steam displacement, not-gas pyrolysis, reverse and forward combustion) were evaluated to select the most effective oil recovery process for the in situ development of the tar sand resource at Tar Sand Triangle, Utah. Reverse combustion was not considered an effective process for this deposit because of the low initial bitumen content and the nonutilization of the residual coke. Since the hot-gas pyrolysis process simulates the pyrolysis of bitumen in the forward combustion process, only steam displacement and hot-gas pyrolysis were chosen for laboratory evaluation using a tubular reactor. In the laboratory evaluation the residual oil and coke saturation after steam displacement was found to be independent of the initial bitumen content and totaled 29% of pore voluem. Since the initial bitumen saturation wasn't much greater than the residual oil saturation, the oil yield from the steam displacement test was only 9.1 wt % of the initial bitumen. The oil yield from the hot-gas pyrolysis test was 54 wt % of the initial bitumen, which is almost identical to the Fischer assay yield. The residual coke and product gas were 23 and 21 wt % of the initial bitumen, respectively. These values are much different from those determined by Fischer assay because of the steam-char and water-gas shift reactions of the hot-gas pyrolysis test. Since the forward combustion process utilizes the residual coke and is expected to produce oil yields equivalent to the hot-gas pyrolysis process, it was selected as the most promising process for the recovery of oil from the Tar Sand Triangle resource. 16 refs., 2 figs., 6 tabs.