this paper models the course tmebling problem as a set of tuples each of which consists of four entities , i.e. lecturers, courses roooms and time- slots that have to be matched in order to construct yhe preferences and their expratie as the constraint of the problem. a bipartite graph that conects the four entities is used as the path of ant movement. in this expriment two metaheuristics algorithm, genetic and ant system algorithm is applied to the problem, and the seults ar compared.