The profile temperature monitoring during aluminum extrusion is a an important information performed in order to avoid profile defects such as burn and crack, or an incorrect tempering, or to preserve the die life and to increase the process productivity. In the last decade die cooling systems through liquid nitrogen have been installed on several extrusion plants. However, a comprehensive and systematic assessment of the liquid nitrogen die cooling effect on process parameters is still missing in literature. In this context, aim of the present work is to develop an advanced numerical finite element model of the extrusion that accounts for liquid nitrogen cooling. The numerical model is then validated over ICEB 2011 Industrial Benchmark experimental trials, where the cooling efficiency in a multi-hollow industrial profile was evaluated. The results of the simulations have been compared to experimental investigation both in cooled and uncooled conditions. The assessment showed a good experimental-numerical agreement in terms of temperature map and extrusion load, thus suggesting the reliability of the novel modelling to support the process optimization.