This paper presents a framework for estimating the time-dependent reliability of prestressed concrete (PC) bridge girders in a marine environment. The analysis considers the spatial variability associated with concrete cover, surface chloride concentration, and coefficient of diffusion of chloride over the entire area. PC bridge girder is modeled as one-dimension spatial model. In addition, random variables for estimating the corrosion process of PC bridge girders are updated based on observational information such as chloride concentration distribution by coring test. Sequential Monte Carlo Simulation (SMCS) is used to update simultaneously multiple random variables relating to observational information. In an illustrative example, the effect of the spatial interval of observational information given by coring test on the updated estimate of a PC bridge girder reliability is investigated.