In this paper, a methodology for quantifying lifetime risk of bridge superstructures is presented. The risk is quantified in terms of the expected direct and indirect losses. A scenario-based approach, which uses the Pontis element condition rating system, is used for identifying the expected losses. The deterioration process of bridge components regarding the transition between the condition states is modeled as a Markov process. In addition, a reliability-based approach is used to compute the probabilities of component and system failure, given the condition states. The methodology is illustrated on an existing bridge in Wisconsin. The expected losses associated with the flexural failure of girders are quantified in time. Furthermore, the time-variant risk-based robustness index is assessed.