A new numerical model for the prediction of temperature development in young concrete structures is briefly presented. With the pre-program. adiabatic hydration curves, which are used to determine the internal heat generation, are calculated. An artificial neural networks approach is used for this purpose. Adiabatic hydration curves, which were included in the learning set, were determined by our own experiments, using the adiabatic calorimeter which uses air as the coupling media. The main program is implemented in the finite element code. This program allows concrete structure designers and contractors to quantify and evaluate the effects of some concrete initial parameters on the adiabatic hydration curves and corresponding temperature development at an arbitrary point in the concrete element. Some examples are also presented and discussed. (C) 2009 Elsevier B.V. All rights reserved