Relating to material properties is a fact that they can vary significantly with temperature. So, this variation mustbe considered in the simulation but to obtain the necessary data is not a trivial task. The measurement of materialproperties, especially at high temperatures, is not only time consuming and costly but also very complex toachieve. Moreover, metal casting industry must also tackle with the usual variations in alloy compositions and withthe evolution of the melt alloy, depending for example of the time in furnace. In case of investment casting anothersource of important variations is the mould. Due to the manufacturing method, it is not possible to assure acomplete homogeneity along the whole ceramic shell. The typical solution consists in the use of material propertiesextracted from literature and handbooks. But they use to be standard values and not always correspond with theproperties of the specific materials used at foundry.The use of inverse modelling techniques is a helpful option to try to avoid these difficulties. These techniquespermit to estimate some model parameters based on direct experimental measurements of related variables. In thecase of metal casting simulation, instead of performing direct measurements of material properties or boundaryconditions, experimental thermal histories are registered in a number of defined points. These thermal histories areused as base and some simulation parameters are modified until reaching a good correlation between numericalsimulated cooling curves and they registered experimentally. This is the approach followed in this work to performthe adjustment between simulation models and experimental results obtained industrially at foundry.