This paper proposes a multi-objective decision model for tamping maintenance plan of ballast track with track geometry state and maintenance costs as sub-objective functions. The optimal sub-objective function of track geometry state was established based on the deterioration model and the recovery model of track vertical irregularities. The minimum sub -objective function of maintenance costs includes planned maintenance costs, fixed maintenance costs, post maintenance costs and comfort loss costs. The improved NSGA-II algorithm was used to solve the model and a corresponding MATLAB program was developed for this purpose. A section of Shanghai-Kunming up line was selected to verify the model and algorithm. Results showed that the proposed multi-objective decision model for tamping maintenance plan of ballast track can provide a number of different maintenance plans, and priority maintenance was given to the section with large initial vertical irregularity and fast deterioration of the standard deviation of vertical irregularity, the maintenance was arranged in the later pe- riod of the annual plan, or no maintenance was arranged for the section with small initial vertical irregularity and low deterioration of the standard deviation of vertical irregularity.