Within the scope of this master thesis a real permutation flow shop problem is solved and optimized by metaheuristics. In particular we implemented hybrid genetic and evolutionary algorithms, ant colony optimization algorithms, simulated annealing as well as iterated local search. These algorithms have been implemented and optimized including customer specific constraints, whereas the final end user, who is represented by a production planner, has the opportunity to choose the constraints under which the problem has to be optimized. First the behaviour of simulated annealing was considered, depending on simulation runs to determine the convergence of the algorithms on several datasets. The algorithms themselves were compared under the same constraints and similar conditions. This permutation flow shop problem comprises to some extent complex constraints, which restrict the solution space but also enable the algorithm to solve certain parts within the whole optimization problem quite easily within short time. The solution space is not homogenous, because we have to consider campaigns which extremely differ in size and restriction of the solution space by their constraints. Next to a detailed problem characterization the metaheuristics are described in detail within the first five chapters of this master thesis to give the reader an overview and an outline of the main ideas of these methods. The last three chapters describe the implementation and discuss the performance of the algorithms and compare their usefulness for the optimization of this real permutation flow shop problem. Within the last chapter we give a recommendation for two algorithms which have delivered the best overall performance on this concrete permutation flow shop problem.