Abstract: More and more stringent environmental regulation to lower the sulfur level in diesel fuel leads to extensive studies on the deep hydrodesulfurization (HDS), which is the key to produce the high quality fuel with low sulfur content. Kinetics and novel rector technologies for HDS have attracted wide attention of both researchers and refineries. In this paper, the reaction network for hydrodesulfurization of dibenzothiophen (DBT) and 4,6-dimethyldibenzothiophen (4,6-DMDBT), in which the sulfur is removed most difficultly, is introduced over various catalysts. The effect of component and support of the catalyst on the ratio of the rate of direct desulfurization (DDS) to that of hydrogenation followed desulfurization (HYD) is analyzed . Various kinetics, such as for model reactants related to DBT with or without inhibitors like H2S, and for the real oil feed, are reviewed in detail. The application of artificial neural network to the simulation and prediction of HDS for various diesel oils is presented. New reactor systems for HDS, such as cocurrent countercurrent trickle-bed reactor, two-phase reactor, catalytic distillation HDS reactor and so on, are summarized. Further researches and challenges for the kinetics of HDS and reactors are proposed.