The growing number of vehicles and pedestrians in the cities increases difficulties for traffic control. Nowadays application of computational techniques helps traffic engineers, giving more efficiency, security and agility to decision making tasks. In this context, a good traffic lights management improves vehicles flow, decreasing traffic jams and delays. This paper applies a multiobjective technique to traffic lights timing using Non Dominated Sorting Genetic Algorithm (NSGA-II), aggregates the proposed algorithm to a microregion traffic simulator coupled into Geographic Information System (GIS). Experiments performed in Porto Alegre area (Brazil) validate the applied technique, comparing obtained results with similar ones from earlier reports.