Due to the increased life of marine structures in the world, repair and maintenance of low cost and high reliability are of concern in this industry. In this paper, the corrosion-induced crack width of reinforced concrete (RC) structure in the marine environment is estimated according to the experimental data, laboratory results, and numerical formulas. Symptom-based Weibull function distribution and Bayesian theory have been used to predict the structural reliability in different modes (without repair and maintenance, with maintenance, and with repair and maintenance). Bi-objective optimization was utilized to provide the client with optimal decisions regarding the probability of structural failure as well as the type of repair and maintenance strategies. The results showed the effectiveness of the proposed method in predicting performance, improving life service, and optimizing repair and maintenance costs of the marine structure.