The kinetics of oxide formation in the presence of water vapour are discussed and compared with oxidation in dry atmospheres. The main protective oxide systems are considered, i.e. alumina, chromia, silica, titania and iron and nickel oxides, and with the possible exceptions of alumina and nickel oxide, oxidation rates are increased by the presence of water vapour. Scale morphology is also influenced by water vapour, and an important observation is that whisker formation is encouraged; this is believed to be due to the more rapid dissociation of water vapour compared to oxygen. In general, water vapour promotes the formation of a more porous scale. This is related to an increase in cation diffusion and consequent vacancy condensation, thereby developing a porous structure. The thermochemistry of oxide formation is discussed, and here oxide stability and hydroxide formation are considered. A significant observation is that where hydroxides or oxyhydroxides form, they generally have higher volatility than the corresponding oxide, and this leads to loss of protection. The effect of water vapour on oxide growth processes is considered. It is demonstrated that all aspects of oxide growth including adsorption, dissociation and diffusion of reactants are altered in the presence of water vapour compared with similar processes in dry conditions. The important first stages of the reaction involving adsorption and dissociation are controlled by the catalytic activity and acid base nature of oxides. For oxides formed at high temperatures very limited information is available, but, in general, data obtained at room temperature is confirmed and strongly suggests that dissociation of any gas molecule is favoured by defects in the surface. Dissociation of water seems to be more rapid at lower temperatures than, for example, oxygen, but this difference may be less pronounced at higher temperatures. Fast diffusion of water in oxides is possible due to "proton hopping", in which protons localised at oxide ions move by transfer from one oxygen to another. Since the OH- ion concentration is increased there is a resultant increase in cation vacancies, and this, in part, is responsible for the observed increase in oxidation rates. A further factor to consider is the possibility of molecular diffusion, and it has been demonstrated that where pores or voids are present in the scale, and the void contains both hydrogen and water vapour, oxidation of the surface nearest the metal will occur by reaction with water to form new oxide and the reaction product hydrogen, while a reduction reaction occurs at the surface of the void nearest the gas phase to produce water vapour. Thus it can be seen that this process provides for rapid inward diffusion of oxygen while the void gradually moves outwards from the metal/oxide interface to the oxide/gas interface. Finally, the review considers the effect of water vapour on the mechanical properties of the scale. Scale adhesion can be improved (iron oxides) or made worse (alumina and chromia) by the presence of water vapour. It is shown that while there is experimental evidence for altered mechanical behaviour, there is very little data on relevant mechanical properties. It is possible, therefore, that water vapour either alters mechanical properties of some oxides, or, as has been demonstrated, the oxide growth process has been changed. Alternatively, at least for the cases where increased oxidation rates were caused by the presence of water vapour, the observed differences between wet and dry behaviour may simply be a function of the greater scale thickness. A significant effort has been made to develop models that can be used to predict the onset scale spallation observed in industrial boilers using process steam. The further development of these models is strongly dependent upon obtaining relevant input data, and this is considered a major challenge for materials scientists. Some areas for future research are proposed. (C) 2007 Published by Elsevier Ltd.