The surface potentials of a (0 0 1) terminated hematite crystal that was annealed at high-temperature were measured as a function of pH by means of the corresponding single crystal electrode. The surface potential at a given pH did not depend on the electrolyte concentration, and was found to exhibit an inflection point. The shape of the function is in phenomenological agreement with the presence of two distinct surface terminations (O and Fe) that have been previously reported for this surface. Aging of the annealed hematite surface, in aqueous electrolyte medium over 2 weeks, leads to a drastic change in the surface potential pH curve. The surface potential becomes that of the ideal O termination. While the O termination data can be modeled using the MUSIC approach, the initial sample that is expected to correspond to the two-domain surface with O and Fe terminations cannot be described within the MUSIC approach based on previously published surface diffraction data. However, the experimental data fall between the O and Fe termination limiting cases when the point of zero potential is placed at the inflection point. The fact that ... mehra surface with the two terminations cannot be modeled may be attributed to various issues, three of which are discussed: (i) the general difficulty to average the potential arising from both terminations, which furthermore are short-circuited via the crystal, (ii) the difficulty of treating patchwise heterogeneous surfaces in surface complexation models, and (iii) the incapability of surface complexation models in their present form to describe potential gradients within the solid. Conclusively, we interpret our results as a transformation from a bi-domain surface, to a single domain surface over time under conditions where bulk hematite solubility is low. Accordingly, the oxygen terminated domain should be the more stable one at this single crystal surface at our experimental conditions.