Power swing is a balanced phenomenon in a three-phase power system， where accurate and fast parameter estimation of oscillation signal is important for the evaluation and elimination of power swing. This paper presents a method for parameter estimation of power system oscillation signal under power swing based on complex spectral interpolation DFT. The proposed method utilizes the symmetrical characteristics of a three-phase system. A complex exponential with the quadrature components is formed from three-phase real signals by using the Clarke transform. The DFT of the complex exponential is then performed. Moreover， various dynamic parameters are estimated by complex spectral interpolation using two DFT samples with the largest magnitude. The experimental simulation results show that the proposed method can accurately and effectively assess the dynamic parameters of power system under the power swing.