Abstract: Parameter inversion has become an effective method for determining strength and deformation parameters of rockfill and other material parameters for practical projects. All the existing inversion methods use a single objective function in analysis of a dam section or even the entire dam body, neglecting the differences in mechanical properties and characteristics of stress and deformation between dam zones of different materials. This paper develops a multi-objective inversion method for the transient and rheological parameters of rockfill dams on the basis of a non-dominated sorting multi-objective genetic algorithm-II (NSGA-II) and a radial basis function (RBF) neural network. In this method, we consider variations in stress and deformation across different material zones and adopt a different objective function for each zone, so that interaction between material characteristics and deformation in each zone can be taken into account through calculation using NSGA-II. The method has been applied to the Shuibuya concrete face rockfill dam (CFRD) for joint inversion analysis on the transient and rheological parameters of its two dam zones of different rockfill materials, namely major and minor material zones. Comparison with the single-objective inversion method shows that the NSGA-II calculations at all the monitored points agree well with the measured deformations and their variation trends, thus achieving a significant improvement.