The Shannon light-oil (32–34° API gravity) steamflood at Naval Petroleum Reserve No. 3, (NPR-3) Teapot Dome field, Wyoming, was initiated in 1985 as a result of favorable production response to steam preheating in the in situ combustion pilot earned out in the early 1980's. Favorable but mixed success of two 10-acre steamflood pilots led to commercial scale operation in 1987. To date, the steamflood project has steamed 210 acres, 90 acres of which is currently being steamed with three 50 MMBtu/hr steam generators. The performance of the steamflood is reviewed in light of the dominant role that the reservoir's complex geology plays, since both injection and production wells are greatly influenced by the localized, highly heterogeneous geology of the consolidated, tight, faulted, and extensively fractured sandstone composing the Upper and Lower Shannon. Pattern step out is dominated by the geologic conditions that dictate each pattern's operation. Economics, mandates for profitability, and budget constraints have influenced the method by which the steamflood has been managed. Investigations of lower injection rate, variation in steam quality and the associated production response, as well as improvements in heat management techniques, are discussed. Increasing steam quality in pattern 5A led to incremental oil production of 44 B/D and a 2.5 times return on investment on the incremental gas used to fire the steam generator. Optimization of the steamflood depends upon accumulating and using information from pattern-specific operations.