It is possible to use concentrated solar energy for producing zinc by direct high temperature dissociation of ZnO. The purpose of the present paper is to show that the overall efficiency of such a process as well as the quality of the products may be controlled by several chemical and physical factors such as heat and mass transfer efficiencies, hydrodynamics and reactor design. The results of three complementary experimental approaches are reported. They are interpreted in connection with simplified models in order to describe the controlling processes occurring in the high temperature and quench sec tions of the reactor. One of the conclusions is that zinc can be recovered by condensation on the cold walls of a heat exchanger and/or under the form of small dispersed particles. Finally, it is shown that the reactor design must be optimized inside the perspective of a given future use of zinc.