Abstract. Timetables are used in different areas among them timetables for schools or roosters. Because it is quite difficult to draw complex timetables by hand, most timetables are generated aided by computers. In this thesis, a framework to describe and solve timetabling problems has been extended. The given framework already included a genetic algorithm which is suitable to solve timetabling problems. Two new optimisation algorithms, tabu search and simulated annealing, have been implemented and integrated into the framework. Tabu search, simulated annealing and the genetic algorithm have been applied to three real world problems. The results have been compared and the tabu search and simulated annealing algorithm have been evaluated regarding their ability to solve timetabling problems. It has been shown that the tabu search algorithm produces significantly better timetables compared to the genetic algorithm and that tabu search generates its solutions in relatively short time. Simulated Annealing produces even better timetables than tabu search and the genetic algorithm.