Over the last 10 years, the emerging EOR process by injection of low salinity water has been investigated through numerous laboratories and some field tests. Extensive research programs have been launched by different teams, aiming at better understanding the involved mechanisms (clay release, cation exchange). One of the main issues is to explain the observed dispersion in additional oil recoveries. In this study, we integrate new experimental observations to the literature to carry out a mechanism analysis of recivery by low salinity injection. These new results were obtained on outcrop sandstones with 9–10 % clay content (Cissokho 2009) and on reservoir cores with reservoir oil.Very positive results to dilute brine injection on the outcrop rock / crude oil of field A system have been observed either in tertiary mode or secondary mode.Corefloods on the same outcrop rock/ crude oil system show that the benefits of low salinity brine injection may vary against the flood temperature.Corefloods in pseudo-reservoir conditions using reservoir cores and the same crude oil of field A showed no response to injection of low salinity brine.Corefloods using reservoir cores and oils from three different fields did not show any positive response to dilute brine injection. In this paper, we analyze the different theories proposed in the literature and present experimental counter-examples for most of them, particularly related to the presence of kaolinite, divalent ions in injected brine, or the effect of temperature. We show that low salinity water injection as an EOR method appears very sensitive to a combination of several parameters. Introduction Low salinity injection is an emerging EOR process as it was put in evidence only 10 years ago by Tang & Morrow (1998). Microscopic efficiency was proved both in secondary and tertiary injection. However, a large scatter in additional oil recoveries was observed and the involved mechanisms are not clearly identified. For this reason, extensive research programs have been launched in the oil industry to understand the underlying physical phenomena: Webb et al. (2004), Mc Guire et al. (2005), Lager et al. (2006). Thus, over the past 10 years, different mechanisms have been put forward in the literature to explain the complex crude oil/ brine/ rock interactions leading to additional oil recoveries when injecting low salinity water. Tang & Morrow (1998) proposed a theory based on the release of mixed wet clay particles from pores allowing the production of oil droplets adsorbed on these clays. This suggests that the system would evolve towards a more water wet state. This phenomenon of fines migration during low salinity injection is well-known and explained by DLVO (Deryaguin-Landau-Verwey-Overbeek) theory of colloids. It is also associated with a permeability reduction resulting from pore throats and pore constrictions plugging by fine mobilization with flowing fluid. Even if several experiment observations are in agreement with this interpretation, contradictory results with additional oil recovery without permeability reduction and no fine production can be listed. Mc Guire et al. (2005) considered a mechanism in which low salinity injection acts like an alkaline solution. The rise in pH is induced by calcite dissolution and cation exchange when dilute brine is injected. It generates in situ-surfactants from crude oil, lowering IFT and then improving oil recovery. This explanation can also be revised by several experimental counter-examples published in the literature.