In systems that operate in the creep range, such as steam power plants, the life-time assessment of highly loaded high-temperature components poses an important task. The main problem in this context is the reliable detection and evaluation of specific material characteristics. First of all there are the strength properties that are the result of the multidimensional interdependences between the individual elements of the chemical composition, the heat treatment parameters and the production conditions. With the current state of knowledge and technology, melt-specific creep rupture strength can only be determined experimentally. Modeling with neural network techniques therefore represents an alternative to analytical methods since multidimensional relationships can be taken into account. This work aims to identify and assess the potential for the application of artificial neural networks to the determination of relevant properties of selected high-temperature resistant steels. The emphasis of the study is to determinate the position of the specific melts in the scatter band of creep rupture data as well as to assess/predict time-to-rupture for the given steel under consideration of all relevant technical data available and to find out an optimum of the creep rupture strength.