Abstract Waterflooding in viscous oilfields poses several problems to both the reservoir engineer and the production engineer. Despite unfavourable mobility ratios, it has been noted that injection performance often exceeds theoretically expected values. In the particular case of the waterflooding pilot project across the Cerro Fortunoso field (Malargue, Argentina), the occurrence of several water supplies of widely varying salinity coupled with the usual problems. This paper describes a pilot study designed to determine the optimal salinity of water to be injected, in order to avoid formation damage resulting from clay deflocculation through what is known in the literature as saline shock. As is clearly shown by the results of the study, abrupt changes in the salinity of injected water result in significant decreases in permeability. According to the study, the problem is minimized over certain ranges. Introduction The Cerro Fortunoso field is located in southern Mendoza, Argentina, in an environmental reservation area known as La Payunia, 135 km away from the capital city of Malargue Department (Figure 1). The field is characterized by its complex structure, which consists of a NNE-SSW striking narrow, elongated, and faulted anticline. Four major east dipping, NNE striking thrust faults are predominant. The oil bearing region belongs to the Neuquen Group, and develops along the anticline flanks. A gas cap, consisting mainly of CO2 (85 to 95% molar content), occurs in the upper part of the structure. Geologically, the field involves fluvial deposits exhibiting considerably marked tectonics with high dips (40 ° to 70 °) and intense NE-SW faulting, giving rise to compartmentalized lenticular reservoirs of poor areal continuity. These lenses are sand/shale intercalations with an average porosity in the order of 18% and mean permeability in the 50 mD range. Background There was concern about the feasibility of a secondary recovery project in the Cerro Fortunoso field, which contains a predominantly asphaltene-based, very viscous heavy oil (14 ° API). Complete information regarding the field's main features and its fluids can be found elsewhere(1). Numerical simulation of the reservoir characterization study performed in Canada with Teknica Overseas Ltd. made comple completion of the project feasible(2). The study warned about the potential negative effects of fresh water injection, due to the occurrence of smectite, illite and kaolinite. At an early stage, the most suitable area for implementation of a pilot project was studied and steps were taken to identify for potential water sources. Thus, we faced one of the main difficulties to overcome. As was described above, the field is located in an extremely desertlike environmental protection area where the nearest river (Rio Grande) is 70 km away. No lacustrine deposits or natural springs occur in the area. Water Sources The use of battery-gathered production water was initially ruled out due to low production rates, high emulsion content, and the significant treatment cost that it would require. The use of two additional water sources was also discussed, i.e., production water separated in the Los Cavaos field and production water from a nearby field which is operated by third parties.