Aerodynamic damping is an important parameter for assessing the wind–structure interaction of super high-rise buildings. A new analysis method is proposed, which first uses the complex blind source separation technique to achieve the effective decoupling of complicated coupled signals and then the Bayesian spectral density approach (BSDA) considering the actual structural aerodynamic characteristics, namely, new Bayesian spectral density approach (NewBSDA), to conduct the parameter identification of decoupled signals effectively. The proposed method is applied to the modal parameter identification of the aeroelastic model test of an 838 m super high-rise building. Results show that the complex blind source separation technique can effectively decouple the coupled signals generated by similar sway modal frequencies in the aeroelastic model. Results of the proposed method are considerably more reasonable than those of the traditional BSDA due to the actual aerodynamic characteristics considered. Comparison shows that BSDA substantially underestimates the cross-wind aerodynamic damping of the structure near the critical wind speed. Given that along-wind aerodynamic loads are wide-band excitations, the aerodynamic damping ratios identified by BSDA and NewBSDA are close. Furthermore, the characteristics of aerodynamic forces identified by the proposed method are satisfactory, which further illustrates the rationality and validity of the method.