The present work aims to contribute to bridge management systems by predicting bridge maintenance necessity. The proposed approach is intended to forecast structural degradation over time. It also aims to predict when corrective maintenance is required. The methodology is based geometric input data and material information of simply supported reinforced concrete bridges (e.g., span length, bridge width, supporting beam dimensions, etc.). Such analysis incorporates empirical and theoretical models of steel reinforcement. A database built from the results of finite element models of bridges was created by varying geometrical and material parameters. The database, consisting of input data and the structural response to standard loads, was used to train, validate and test an Artificial Neural Network (ANN). The database was expanded until the ANN was reliable enough to forecast the structural response of a bridge not included in its training input set. Then, a MATLAB routine was developed to provide, for any simply supported reinforced concrete bridge within the established limits of geometric and materials parameters and by using simple environmental parameters, the expected period of time until the main degradation events due to corrosion. The results of multiple simulations show that water-cement ratio, concrete cover and concrete compressive strength have significant impact on the lifecycle of concrete bridges under corrosion attack. On the other hand, the distance from the sea coast is less impactful.