The injection of water with alteration of salinity is an advanced recovery method that has been promising over the years, as it stands out due to its economy and simple application, which significantly increase the recovery factor of reservoirs when compared with the other techniques used. For carbonate reservoirs, defined by having complex characteristics and preferential wettability to oil, this technique is even more attractive, resulting in alteration of wettability, an important factor in increasing oil recovery. The literature highlights the influence of the composition of the injection water; however, the physicochemical mechanisms involved have not been fully understood and so far generated discussions. This research aims to evaluate the recovery of oil in carbonate reservoirs by reducing the concentration of sodium chloride in the injection water. Coreflooding tests were conducted, with experimental conditions of 8230 psi and 69° C, representing the Brazilian Pre-Salt carbonate reservoir. Four rock samples were used. The tests were divided into three series; each series is composed of three tests. Three different brines concentrations are used in each tests. The injection waters analyzed were: formation water (FW), seawater (SW), desulfated seawater (SWDs), seawater depleted in NaCl (SW0NaCl), desulfated seawater depleted in NaCl (SWDs0NaCl), seawater depleted in NaCl with the addition of sulfate ion (SW0NaCl+SO 2− ) and desulfated seawater with a 50% reduction of NaCl (SWDs0.5NaCl). In addition to the brines used, a mixture of oil with a simplified gas composition was used to represent the fluid condition in the reservoir. The results obtained demonstrated that the most promising injection waters were: SWDs0NaCl (tertiary recovery), which resulted in total increases of 13.5% in production and SW0NaCl (final recovery), which resulted in total increments of 14.8% in more than four porous volume injected. The influence of the sulfate ion could also be observed. SW0NaCl+SO 2− was not effective, which may indicate that the proposal to increase the amount of sulfate in the last injection water is not valid for this type of lithology and operational conditions. Regarding the analysis of water used as the first injected water (FW, SW and SW with change in salinity), all proved to be efficient, the FW and SW resulted in tertiary recovery. However, it was possible to observe that the amount of oil produced with few porous injected volumes of FW is much greater than with SW, which indicates that oil production is faster with the injection of formation water, which fluid is not considered reactive. Concerning to SW with change in salinity, it recovered larger volumes of oil compared to other waters in this mode. In conclusion, the low salinity water injection is efficiently increases oil recovery in carbonate reservoirs.