Oxide layers generated on the surface of metallic structures may play a protective role beneficial or otherwise undesirable (mill rolls for example).They can generate irreversible damage to the structures but they may also provide protection of the base metal against environmental conditions.Thus, depending on the application, one has to promote both the development these oxide layers and their separation.It is therefore important to develop predictive models to describe the mechanical states developed during the growth of oxide layers and due to loading that can undergo oxidized structures.Conditions of service strength depend strongly on the conditions of implementation of the oxide layers and one is faced with a problem of parameter optimization taking into account the history of thermal conditions (temperature rise and cooling, weather,conditions...) for a prescribed performance, while minimizing the cost of implementation of the layers.We present an approach based on theoretical developments, experimental and numerical prediction and identification permitting the identification of the properties of metal/oxide systems.Applications of the approach on systems Ni/NiO and Zr/ZrO2 illustrate this approach.