By establishing the quantitative relation between condition monitoring of overhead lines and its failure rate，a multi-objective two-layer optimization model for selecting overhead lines with minimal maintenance cost and system risk is developed.To prevent the conventional transformation of multi-objective optimization into single-objective optimization，the non-dominated sorting genetic algorithm-Ⅱ（NSGA-Ⅱ）is adopted to directly obtain the non-inferior solution set of the two-layer optimization model.By making use of the relation between the two objectives of maintenance cost and system risk reflected by the non-inferior set，the to-be-maintained overhead line can be determined provided the risk is acceptable.The overhead line set for maintenance is directly obtained from the non-inferior solutions representing the relation between the two objectives of maintenance cost and system risk by using an acceptable system risk level.The maintenance strategy determines which lines should be maintained so that the system reliability can be maintained at the acceptable level at the lowest cost.A case study using the RBTS system demonstrates that the method proposed can avoid the disadvantage of over-arranging the maintenance lines in the traditional condition based maintenance.