Noting the importance of flexible optimization for methods managing conflicts resolution of complicated and manifold problems, in this paper, we study mixed-integer programming problems(MIP) under multi-objectives. As known from the term NIMBY (Not In My Back Yard), site location problems of hazardous wastes are eligible case studies in such a situation, that is to say, associated with human, environmental and economic concerns. Showing that general formulation of such site location problems refers to MIP under multi-objective, we preliminarily discuss how to apply a genetic algorithm (GA) as a practical and effective solution method for MIP. Then we have proposed a hierarchical approach named hybrid genetic algorithm (HybGA) which is characterized by the combined use oi genetic algorithm with appropriate mathematical programming. Mentioning about some promising features of the idea, we extend it to solve multi-objective mixed-integer programs, and reveal the superiority of HybGA over the conventional multiobjective methods of GA. Finally by taking a site location problem of hazardous waste disposal, we have examined numerically the effectiveness of the proposed approach through comparison both with MOGA and branch and bound methods.