Various diffusion experiments using geologic media have been carried out and it is often assumed that aqueous diffusion is the dominant transport mechanism. However, in some cases diffusive migration has been much faster than predicted in the model simulation. To explain such results surface diffusion of sorbing species was invoked. Experimental results were generally open to interpretation but possible existence of surface diffusion, whereby sorbed radionuclides could potentially migrate at much enhanced rates, necessitated investigation. The potential for surface diffusion of some sorbing nuclides on through-diffusion experiments using domestic rocks was examined. The apparent diffusion coefficients for sorbing cations were determined from their steady-state diffusion flux through rocks disks, while effective and pore diffusion coefficients were obtained with non-sorbing tracers through the same rocks. Diffusive transport models through domestic granites and granodiorites based only on pore diffusion did not often described adequately for sorbing cations. Thus, surface diffusion should be considered. Then what was the most important measure to estimate surface diffusion? As far as we examine, the sorption reversibility provides a hint of surface diffusion. The reversible sorbing species, for example, Sr, has a remarkable surface diffusion contribution, whereas surface diffusion has a relatively small contribution for irreversibly sorbing species such as Cs and Am under domestic experimental conditions.