Reservoirs containing very high total dissolved solids and high hardness make the design of a surfactant polymer (SP) flood extremely difficult because surfactant tends to precipitate and separate under these conditions. Beside divalent ions, Ca2+, Mg2+, presence of iron in the brine can be a challenging issue. Different surfactant formulations are evaluated and incorporate cosurfactants and co-solvents which minimize viscous macroemulsions, promote rapid coalescence under Winsor Type III conditions, and stabilize the chemical solution by reducing precipitation and phase separation. The optimal surfactant formulations were further evaluated in one-dimensional sand packs and coreflood tests using Berea sandstone, reservoir oils, and brines at reservoir temperatures. Using similar injection protocols, 0.5 pore volumes of surfactant/polymer + 0.5 pore volumes of polymer drive, experimental results show the oil recovery ranging from 46 % to 89% of the residual oil (Sor) after water flooding. The level of surfactant loading is less than 0.5 wt%. A single-well test was conducted to confirm laboratory results in situ in the presence of high-salinity formation water containing 185,000 mg/L total dissolved solids (TDS). The test is considered to be a technical success and confirms the effectiveness of the high-salinity surfactant-polymer formulation (0.4 wt% of surfactant and 1,800 ppm of polymer loading). The Sor was reduced from 25% to 7% resulting in approximately 72 % of the residual oil being mobilized. A pilot test at the same reservoir is scheduled to be performed in 2012 to further evaluate the effectiveness of surfactant formulation and address technical issues related to scale-up.