A one-dimensional mathematical model for the corrosion of a metal surface under a porous layer (often an oxide layer) is presented using porous electrode theory. The model allows one to predict the rates of corrosion in a metallic system where oxygen reduction is the only cathodic reaction which occurs on the surfaces of both the metal and the porous oxide layer albeit at different rates. Thus the model simulates the scenario where the porous layer residing on the metal surface competes with or complements the oxygen reduction reaction happening on the metal surface. The study finds that a large thickness, low porosity, low oxide electrical conductivity, poor oxygen reduction on the porous layer, and high specific contact resistivity of the metal–oxide lead to poor spatial separation of the anodic and cathodic reactions and hence encourage self-repair. In particular the metal–oxide contact potential and the oxygen reduction rates on the oxide surface are shown to have profound implications for self-repair strategies for corrosion mitigation.