Many decision problems in the real life are multi-objective ones. The natural characteristics of multi-objective decision are: there are more than one objectives, and the estates of every objective is not in the same and there's contlict between them. Therefore, multi-objective optimization problems (MOPs) is always the hot and hard point of optimization. The aim of MOPs is to generate a list of solutions for problems called the pareto set. Evolutionary algorithms can efficiently solve MOPs by obtaining diverse and near-optimal solution sets. In this paper, the author present methods to combine Genetic Algorithm and Tabu Search to solve multi-objective otimization problems. The result of those methods was verified by testing some concrete problems.