In this paper two models are presented for a three-phase catalytic packed bed reactor in which in evaporating solvent is used to absorb and remove most of the reaction heat. A plug flow model and a model comprising mass and heat dispersion in the reactor are discussed. The results of both models are compared to each other and to experimental data obtained in a miniplant on the hydrogenation of 2,4,6-trinitrotoluene to triaminotoluene described in Part I. The influence of reactor pressure, feed temperature and the molar ratio of hydrogen to the reactant in the feed are discussed. It is concluded that both models can well describe the influence of the operating variables on the reactor behaviour and that a large part of the reaction heat can be removed by evaporation of the solvent. From a comparison with the experimental results it is concluded that the dispersion model can well describe the experimental data for a fresh catalyst. The agreement between the model and the experimental data for a deactivated catalyst is not good. As long as no quantitative description of the local deactivation in the packed bed is available, any model will fail to predict local concentrations in the reactor. Despite this drawback recommendations can be given how to use this reactor type of optimize the selectivity in producing an intermediate product.