Melt growth processes provide the basic crystalline materials for many applications. The research and development of crystal growth processes is therefore driven by the demands which arise from these specific applications; however, common goals include an increased Uniformity of the relevant crystal properties at the micro- and macro-scale, a decrease of deleterious crystal defects, and all increase of crystal dimensions. As melt growth equipment and experimentation becomes more and more expensive, little room remains for improvements by trial and error procedures. A more successful strategy is to optimize the crystal growth process by a combined use of experimental process analysis and computer modeling. This will be demonstrated in this paper by several examples from the bulk growth of silicon, gallium arsenide. indium phosphide, and calcium fluoride. These examples also involve the most important melt growth techniques, crystal Pulling (Czochralski methods) and vertical gradient freeze (Bridgman-type methods). The power and success of the above optimization strategy, however, is not limited only to the given examples but can be generalized and applied to many types of bulk crystal growth. (C) 2002 Elsevier Science B.V. All rights reserved.