It is generally accepted that a buried steel structure cannot be cathodically polarized to a potential much more electronegative than approximately -1.2 VCSE, due to the limiting potential encountered with the electrolysis of water. Nevertheless, cathodic protection data obtained in the field often includes of-potential measurements significantly more electronegative than would seem possible in theory. While there are a variety of possible explanations for such erroneous data (e.g. the failure to interrupt all influential current sources), it is proposed that in some cases, the polarization of the soil itself may be responsible. Time-domain induced polarization (IP) is a well-established geophysical survey technique in which the polarization characteristics of the soil are quantified, as a means of identifying mineral deposits. This paper discusses the theory behind the IP technique, and suggests how the polarization of certain soils might lead to off-potential measurements which are more electronegative than the true polarized potentials of a cathodically protected structure. Off-potential data from one case history is presented to sup-port this theory.