A one-dimensional home-made math. model of coke-making process in the heat-recovery coke oven was developed to predict behavior of the oven and to optimize its design as well as operation. The model can be applied to improve quality and homogeneity of the coke produced. An inverse procedure was used in this model to represent carbonization of particular charge and is based on temp. measurements within the coal coke bed only. The model provided some information not available usually during coke making. The data collected during commissioning were used to validate the developed algorithm. The temp., pressure and gas compn. were measured in several selected locations within the industrial oven.