This paper takes the multiobjective traveling salesman problem (MOTSP) as the representative for multiobjective combinatorial problems and develop a set-based comprehensive learning particle swarm optimization (S-CLPSO) with decomposition for solving MOTSP. The main idea is to take advantages of both the multiobjective evolutionary algorithm based on decomposition (MOEA/D) framework and our previously proposed S-CLPSO method for discrete optimization. Consistent to MOEA/D, a multiobjective problem is decomposed into a set of subproblems, each of which is represented as a weight vector and solved by a particle. Thus the objective vector of a solution or the cost vector between two cities will be transformed into real fitness to be used in S-CLPSO for the exemplar construction, the heuristic information generation and the update of pBest. To validate the proposed method, experiments based on TSPLIB benchmark are conducted and the results indicate that the proposed algorithm can improve the solution quality to some degree.