This paper provides a discussion on the main changes required in order to extend particle swarm optimization to the solution of multi-objective optimization problems. A short discussion of some potential paths for future research in this area is also included.