Although the design life for bridges is mandated in AAHSTO LRFD to be 75 years, the observed service life is shorter due to the deterioration of various elements over the years. The major reasons of the deterioration can be attributed to two causes: 1) rapid increase of the live load (weight and frequency) and the gradual loss in the load carrying capacity. The decrease in the load carrying capacity is caused by the deterioration of material and environmental attacks. Therefore, in order to quantify the deterioration of the structural performance and implement appropriate maintenance and repair schemes, it is important to forecast the long-term performance of the bridge. In this study, the research is aimed at developing a probabilistic approach to forecast the long-term performance of the bridge. A prediction model is developed for live load based on 20 years of Weigh-in-Motion (WIM) data collected from various sites in New Jersey. Based on the historic inspection records and probabilistic techniques, including Monte Carlo simulation, the deterioration models are developed for various structural members including prestressed concrete girder, reinforced concrete deck and steel girder. Using the developed live load and deterioration models, the long-term performance of the bridge under various conditions is forecasted and the recommendations are provided for future application of proposed approach. It is predicted the performance of the prestressed concrete girder and steel girder deteriorates around 20% to 26% over 75 years while the deck would be needed for replacement over 40 years under severe environmental condition.