Prestressed concrete bridges that were erected during the mid-20th century, representing a significant part of German infrastructure constructions, are not only approaching the end of their intended service life. They are also facing accelerating degradation due to increased live loads, climate-induced environmental stressors such as more extreme temperature cycles, increased precipitation, and elevated humidity. These evolving conditions exacerbate concrete deterioration. In addition, stress-corrosion cracking of internal prestressing tendons undermines the residual load-bearing capacity of the structures. Conventional non-destructive testing methods remain largely incapable of detecting internal tendon damage reliably. To address these challenges, an inverse system-identification framework that integrates a high-fidelity finite-element (FE) model with adjoint-based sensitivity analysis for the localization and quantification of tendon and concrete damage is implemented. The FE model is parameterized elementwise by the Young’s modulus of concrete and the steel cross-sectional area of tendon elements. The fine spatial resolution of the mesh yields a large parameter set, which is estimated via a local, gradient-based optimization algorithm. Gradients of the discrepancy between simulated and measured structural responses are computed efficiently via the adjoint method, enabling the solution of the minimization problem. Due to practical limitations, the number of measurement locations is limited. Applying sensors to real world structures results in large manual effort. By utilizing only a limited number of displacement or strain measurements, this approach recovers the spatial distribution of concrete stiffness and tendon cross-sections, pinpointing damage locations. This contribution focuses on a number of case studies, showing the feasibility of this approach and the transferability to real world structures.