Considering the increasing number of structures constructed before decades, it is important to assess the residual seismic performance Considering the increasing number of structures constructed before decades, it is important to assess the residual seismic performance networks. In this study, the model parameter updating problem of a seismic-isolated bridge pier with the deterioration in both the rubber bearings and reinforced concrete pier is investigated using the simulated seismic response data. We employ a well-known Bayesian inference algorithm, called transitional Markov chain Monte Carlo, with a modification reducing the correlation between samples in Markov chains for accurate and efficient estimate of the updated model parameter values. This application clarifies the effect of the noise level of the observed data as well as the periodic characteristic and intensity of the input ground motions to the accuracy of the updating results.