The measured stresses associated with the growth of oxide on the surface of aluminum are much lower than those calculated from the Pilling‐Bedworth ratio. The magnitude of the stresses in alumina formed anodically on pure aluminum is shown to be dependent on the rate of formation. Some experiments are described which show that, even at low temperatures, the presence of a large ionic flux will permit the deformation of alumina providing a mechanism by which the growth stresses can be relieved.