The application of optimization techniques to real-world problems demands for efficient modeling and encoding of the problems as well as for the application of fast and feasible algorithms. Many practical problems have a multiobjective character. In general, seen from a mathematical as well as from an empirical point of view, this class of problems is extremely complex. An engineer has to face and master multiobjective problems efficiently every day. All the more important is a systematic and compact summary of the state of the art of practical mathematical statements, modeling techniques, optimization algorithms and statistical tests. This thesis gives an extensive and practical overview from theory to various deterministic and stochastic techniques of multiobjective optimization. A special focus lies on multiobjective evolutionary algorithms (MOEA). Practical examples from literature are augmented by a thorough discussion of a real-world structure and design optimization problem - the mold temperature control design problem. Additionally, a side view to parallel multiobjective optimization is given. New algorithms and metrics are introduced such as the parallel pMOHypEA, a MOEA which uses hypergraphs for the structuring of the population, DOPS, an improved multiobjective particle swarm algorithm, or the MMBBH-metric, which is a variant of the attainment surface technique. The application of statistical design of experiments to parameter optimization of multiobjective optimization algorithms illustrates systematic comparison and meta-optimization techniques. The mold temperature control design problem is the exemplary focus of many analyses and, therefore, helps to increase the comprehensibility of the discussions.