<p>This paper presents an experimental study on the Bayesian model updating of a laboratory-scale pedestrian bridge model using complex modal data. Using complex modal data for model updating is crucial as it accounts for non-classical damping, a phenomenon frequently encountered in real-world applications. In this study, to update the FE model, complex modal parameters are identified from the acceleration measurements of the real laboratory structure. Usually, in an experimental environment, modal datasets vary and include non-negligible posterior uncertainty. These posterior statistics are used in addition to the Most Probable Values (MPVs) of natural frequencies, damping ratios, and mode shapes as data for an accurate posterior estimation. To capture modal data, wireless sensors were used to record the response on the desired (as accessible) component of the experimental model. Notably, the use of such a low number of sensors is not accompanied by the need to assemble global mode shapes. The results show a good match between measured and predicted modal parameters with an error percentage less than 5% which validated the updating methodology accuracy. This is further supported by a high correlation in terms of MAC values &gt; 95% between updated and measured mode shapes. In total, the method proved computationally efficient, robust against incomplete measurements, and effective in detecting damage. These results demonstrate that Bayesian updating using complex modal data can be reliably applied to realistic laboratory-scale structures under partial instrumentation.</p>