This work presents a new approach to the ion dynamics in concentrated electrolyte solutions as those observed in natural environments or industrial fluids. The dependence of the ionic transport coefficients as a function of the concentration, has been obtained from the equilibrium distribution functions of the mean spherical approximation (MSA). At first, this work uses a Smoluchowski level theory. Mutual diffusion of dissociated and associated electrolytes have been particularly studied by analytical calculations and Brownian dynamics simulations. The self-consistency of the theory, which is obtained if the same parameters are used for the description of the different transport and equilibrium quantities, is one of the major aim of this thesis. It has been obtained after a careful analysis of the reference-frames of the different transport coefficients, which had not been done previously. The properties of solutions up to a concentration around 2 M can be described by using only the infinite dilute self-diffusion coefficients and the radii of the ions in solution. These results agree with the Onsager limiting laws and generalize the latter. The short time dynamics of electrolyte solutions has also been studied. A mode-coupling theory, in agreement with Brownian Dynamics simulations, allowed to calculate the variation of the self-diffusion coefficients as a function of the characteristic time of the measurement and the con- centration, with the same parameters than those used previously for the transport and equilibrium properties. For short-time observations, the relaxation effect does not have the time to occur : the diffusion process is faster. Thus, mode-coupling theory explains the fact that short-time measurements (QENS) give larger self-diffusion coefficient values than long-time ones (NMR, tracers). The description of electrolytes at charged interfaces, in the case of hydrated clays, has also been revisited in the course of this thesis. The continuous solvent description was found to be consistent with discrete solvent microscopic simulations.
