In natural marine exposures, the chloride profile may have a maximum in the concrete surface (surface concentration) or this maximum may appear some mm or even centimeters, in depth. This fact has been attributed to a) a “washing effect” due the continuous action of the sea water, b) the wet and dry “convection” induced by the action of the sea water waves c) the leaching of the hydroxide ions or d) the carbonation of the external layer of concrete. In a test in the laboratory, the profile resulting in a diffusion chloride test in most of the cases presents the maximum in the surface of the specimen but seldom has this maximum beyond the surface. This behaviour in natural and in laboratory conditions, although noticed, has not been satisfactorily explained. The author has addressed the problem in a previous communication. In present one shows evidences on the role of the carbonates/bicarbonates in the water of the testing solution. It can be deduced from the analysis of the profiles obtained with de-carbonated solutions that the carbonates/bicarbonates ions compete with the chlorides in the binding with the hydrated cement phases, which results in a modification of the chloride profile and a different diffusion coefficient. The maximum seems to be produced by the higher mobility of the chloride with respect to the carbonate/bicarbonate. The bound chlorides are displaced by the carbonation and the free chlorides move inside forward. In the external parts less total chlorides are noticed because the free chlorides are function of the bound chlorides.