In the calculation of chloride diffusivity based on migration testing, the external potential is normally applied without any correction for the potential drop in the system outside the test specimen. In the present paper, it is shown that this potential drop may give an error of up to 20% of the chloride diffusivity obtained based on the nominal (external) applied potential. For different types of concrete (w/c = 0.40–0.60) and different salt solutions (1–3% NaCl in the upstream cell and 0.3m NaOH in the down-stream cell) and different levels of applied voltage (10–60 V), a total potential drop in the migration system varying from approximately 2.5 to 3.0 V was observed. Since measurements of such a potential drop make the routine testing of chloride diffusivity more elaborate, a simple correction factor of 2.5 V for the potential drop in the migration system outside the test specimen may be introduced.