A lot of chemical, petrochemical or refining processes require contact between three phases : a liquid feed, a gaseous reactant and a solid catalyst. Frequently, the catalyst activity is reduced by poisoning of active sites or coke deposits. This is especially the case with the processes used in heavy residual oils hydrotreating. As the catalyst life is reduced, the substitution or regeneration of the inactive catalyst is frequently necessary. Various solutions, such, as fixed beds used with swing reactors, fluidized beds, or moving beds with down flow of the catalyst and co-current or counter-current of the feed, can be proposed to perform this task with a minimum of time and production losses. A theoretical comparison between the performances of the various technologies has been made by means of a detailed simulation of the behaviour of each of these catalytic beds over a long period. Of course, in the models, some assumptions are necessary, like the ideal fluid and solids flows. Nevertheless, the problem remains complex because hydrodynamic, kinetic, catalyst deactivation, or thermal effects occur simultaneously, within the particules and/or in the bed as a whole. Various pilot plant data are of course used in order to build the kinetic part of the models. This comparison shows a marked advantage for the moving bed with counter-current flow between feed and catalyst owing to the systematic optimum use of the catalyst potential. Consequently, a series of experiments was made on various sized cold mockups designed to simulate counter-current movind beds. These experiments were necessary to demonstrate the feasability of the process, to specify the relations among gas and liquid superficial velocities, particles and fluids properties, and hydrodynamic regimes, and to develop the scale-up rules. The main goal is to secure a uniform distribution of the two fluids through out the whole bed of catalyst, and at the same time a regular progression of the solid particles during the sequential withdrawal. It is the reason why the movement of the solid and, moreover, the residence time distribution of the catalyst particles have been especially followed. To this end, a new technique has been developed to measure the concentration of magnetically marked particles. Furthermore, a specific effort has been devoted to technological problems such as selection and developement of special valves able to insure efficient tightness between the various sections at elevated temperatures and pressures in the presence of solid particles. Another study has been devoted to the operation of addition and withdrawal of the catalyst to and from the reactor running at high pressure and temperature. The solid flows and the catalyst transfer conditions have been carefully studied and tested. Finally a demonstration unit has been built at a semi-industrial scale in a mini-refinery. It was a convincing way to check all the conclusions drawed from the developement step with models, smaller scale pilots and mock-ups. As the fluids and the catalyst were real industrial products, the feasibility of the whole process has been demonstrated. Therefore, a new industrial process using the countercurrent moving bed is now available and permits a very efficient use of the catalyst particles.