Multiobjective optimization using evolutionary algorithms (EA) has become a wide area of research during these last years. However, only a few papers deal with the handling of preferences. To take them into account, a method based on multi-criteria decision aid, PROMETHEE II, has been implemented. Further-more, as the handling of the constraints is very critical in EA, an original approach (PAMUC Preferences Applied to MUltiobjectivity and Constraints) has been proposed, which considers the satisfaction of the constraints as another objective.