Finite Element Model Updating (FEMU) plays a critical role in predicting the dynamic behavior of existing structures, particularly when reconstruction alters their original configurations. This study focuses on the comparative performance of five optimization algorithms —'interior-point,’ ‘sqp,’ ‘sqp-legacy,’ ‘active-set,’ and ‘Bayesian optimization'—applied to the FEMU of a Gerber-type bridge in Croatia that was reconstructed into a static system. The research emphasizes the importance of integrating experimentally determined dynamic properties, such as natural frequencies and mode shapes, into numerical models to enhance their predictive capabilities. The initial model was refined through an iterative FEMU process to more accurately represent the actual bridge behavior by adjusting boundary conditions, material properties, and support stiffness. Key differences between the initial and updated models highlight the significant impact of parameter adjustments on dynamic response accuracy. The comparative analysis of the algorithms is conducted based on the discrepancies between the predicted and actual dynamic behaviors, as well as the computational time required for each optimization process. Results show varying levels of accuracy and efficiency among the algorithms, providing valuable insights for selecting appropriate methods in similar structural assessments. This study underscores the importance of using experimental data in the model updating process to reduce uncertainty in structural predictions, and demonstrates the effectiveness of different optimization techniques in achieving more reliable numerical models. The findings offer practical guidance for engineers involved in the FEMU of bridges and other civil structures, promoting enhanced accuracy and efficiency in dynamic behavior predictions.