Segregation is an important phenomenon which heavily affects the final mechanicalproperties of steel products. The presence of several complex physical phenomena resultingin final segregation pattern in as-cast products makes the quantitative prediction ofmacro-segregation for industrially relevant casting processes extremely difficult. In thepresent work, a reliable prediction of important rolled product quality (in terms ofmechanical and Charpy impact properties) which are linked to segregation is achieved forplate rolled pipeline steels by exploiting data related to the as-cast structure andcaster operational data (including casting machine condition) through the application ofneural networks. In particular, a hierarchical approach is proposed for the prediction ofthe Charpy Impact Value, in order to reflect the physical link between this quantity andthe Ultimate Tensile Strength. The neural predictor has been developed by exploiting realindustrial data and its performance can improve through time by enlarging the databasethat is used for its training.