Impact loads applied to concrete sleepers, either due to the wheel flats of a train or rail abnormalities, may cause the sleepers to crack. The rail pads have been used to attenuate the effect of these impact loads on concrete sleepers and on the substructures in ballasted railway tracks. The rail pads experience high intensity loads due to wheel impacts and, as a result, the pad properties deteriorate during the service life. The consequences can be significant for the dynamic behaviour of railway tracks, interaction of vehicle and track, and impact forces on track components. However, there is currently no standard method available that can be used to evaluate the dynamic characteristics of the pads. In this paper, the degradation of rail pad properties as a function of their in-service life is studied with a view of developing a technique for predicting the optimum period of track maintenance with regard to pad replacement. Identification of structural properties using methods of experimental modal analysis provides a means for investigation of the deterioration of dynamic pad characteristics. A technique using an instrumented impact hammer is employed in this study. The pad samples were collected from the railway lines in Sydney Electrified Network, Australia, which are operated by Rail Corporation New South Wales (RailCorp). Through testing of pads in a recently developed rail pad testing machine, the vibration responses were measured in the frequency range from 0 to 1000 Hz using the Bruel & Kjaer PULSE Vibration Analyser system. The analytical solution was derived to estimate the dynamic stiffness and damping constants of the worn pads from the obtained experimental data. The rates of deterioration based on limited data are proposed to predict the useful lifetime of pads and the period of pad replacement. In this study, the type of rail pad used was the high-density polyethylene (HDPE) pad with 5.5 mm thickness. All worn pads were sourced from RailCorp tracks with approximately 22 MGT (million gross tons) annual tonnage. Based on linear regression analysis, it has been determined that at 20 kN preload (equivalent to PANDROL e-Clip fastening system clamping force) degradation of dynamic stiffness and damping is about 2.2 MN/m and 19.6 Ns/m per MGT, respectively, for the particular track conditions.