Many field tests of the steam assisted gravity drainage (SAGD) process have been conducted and have shown that the process is a technically effective one at extracting oil from heavy oil and bitumen reservoirs, However, it has not been firmly established whether the technology is operated at optimized conditions to yield maximum economic returns. This is especially important because typically SAGD depends on the combustion of natural gas to generate steam and this is the dominant cost. The cost of natural gas can be significant when natural gas prices are high. This research evaluates the use of a genetic algorithm optimization scheme to control a commercially available thermal reservoir simulator in order to optimize the steam injection strategy to reduce the cumulative oil to steam ratio (cSOR). The reservoir description is typical of that from a low to medium quality Athabasca reservoir. The results show that the injection strategy can be altered to reduce the cSOR up to 50% from a uniform injection pressure strategy to I after the steam injection strategy has been optimized. The optimized profile has high steam injection pressure at the beginning of the process before the steam chamber reaches the top of the oil-rich zone. Before the chamber reaches the oil pay, with high injection pressure, the saturation temperature is high and there are no thermal losses to the overburden. After the chamber reaches the top of the formation, the injection pressure is lowered throughout the remainder of the process. This reduction of injection pressure implies that the saturation temperature falls and consequently the losses to the overburden are lowered. Thus the overall thermal efficiency of the process is enhanced. The optimized strategy is compared to processes operating at 1,000 and 2,000 kPa constant injection pressure.