The objective of this study was to better understand the recovery of human skin impedance following iontophoresis in vivo. Volunteers were subjected to a 15-min period of iontophoresis in the presence of aqueous solutions of either NaCl, KCl, CaCl(2) or MgCl(2) at 133 mM. Subsequently, the low-frequency impedance (at 1 Hz) recovery was followed for a further 30 min. Assuming direct proportionality between the reciprocal impedance and the ion concentration in the membrane, the experimental data were fitted to the appropriate solutions of Fick's second law of diffusion to derive characteristic diffusion parameters (D/L(2)), apparent diffusivities (D), diffusion pathlengths (L) and mobilities, and ion concentrations in the skin immediately post-iontophoresis. Ion fluxes out of the membrane after termination of current flow were also deduced. In general, recovery was relatively independent of the background electrolyte as previously reported, and the data were consistent with ion transport in predominantly aqueous pathways. Compared to its mobility in aqueous solution, however, the apparent Cl- mobility in the skin was smaller, presumably due to the fact that, under normal physiological conditions, the human skin barrier supports a net negative charge. In parallel, the initial "release" of Na+ and K+ from the skin post-iontophoresis was faster than that of Ca(2+) and Mg(2+), the latter cations of higher charge density being able to associate more strongly, it seems, with the negatively-charged skin. The simple physicochemical analysis of the data presented serves to emphasize that a decrease in skin impedance is not a manifestation of damage to the barrier--rather, it is a natural response to the relevant electrical potential and ion concentration gradients involved.