Abstract: Three damping identification methods based on time response functions are studied in this paper, including the logarithmic decrement method, the Hilbert transform method and the wavelet transform method. Thealgorithms of these methods are presented and the ability of these methods to identify close modes is analyzed. The damping estimation accuracy is calculated using noise contaminated response with 5\%, 10\% and 30\% white noises, respectively. Results show that the wavelet transform method is more robust than the logarithmic decrement method and the Hilbert transform method. The wavelet transform method is then used to identify the modal parameters of Runyang suspension bridge using the vibration responses obtained by Structural HealthMonitoring System. Modal parameters of six low order modes are identified. The difference of the modal frequency between the second mode and the third mode is only 0.015\,Hz. Therefore, the wavelet transform method can identify modal damping of closely spaced modes from noise contaminated data. The method should be a good choice for engineering applications.