AbstractThis article aims to develop a hybrid approach of complementary ensemble empirical mode decomposition (CEEMD) and random decrement technique (RDT) for identifying the modal parameters (i.e., natural frequency and damping ratio) of structures using high-rate (50 Hz) global navigation satellite system and real-time kinematic (GNSS–RTK) measurement data. The framework of the proposed approach included two stages: (1) CEEMD was first used to derive a set of single-component signals, called intrinsic mode functions (IMFs); and (2) RDT was employed to extract the free decaying signals from the IMFs, including the main frequency information of structures. Subsequently, a three-story shear building model and an actual multispan bridge were used to validate the proposed approach. Additionally, a finite-element (FE) model of the bridge was built for comparing with the experimental case. Finally, the results indicated that the GNSS–RTK technique is capable of monitoring the dynamic displacements of multispan bridges with reliable accuracy. The proposed method performed better than the classical natural excitation technique—eigensystem realization algorithm (NExT–ERA) and stochastic subspace identification (SSI) algorithm—and was viable in practical applications.