Based on adaptive genetic algorithm and grey relation degree, a new algorithm for multi-objective decision is put forward. First, the multi-objective problem is divided up into multiple single-objective problems. Then the optimal results of these single-objective problems are respectively solved and unified which consist of referenced vector. Second, the grey relation degree between each individual and the referenced vector is calculated, and the grey relation degree is assigned as the fitness of the individual. At last, the Pareto optimal sets are selected by using adaptive genetic algorithm so that variety of population is saved. Simulation examples show the effectiveness of the proposed approach.