This study aims to report a long-term scour monitoring of an in-service historical masonry arch bridge using microtremor monitoring. Scour, the localized erosion that undermines bridge foundations during flooding, is considered one of the most critical threats to bridges crossing rivers worldwide. In UK and Europe, in-service historical bridges are potentially high risk for scouring due to the escalating demands imposed by the modern transport systems, but the application of structural health monitoring (SHM) is still limited. Measuring the dynamic responses of historical structures during period of swollen river water and analysing structural performances by installing long-term monitoring systems is crucial for ensuring safe operation during flood events. This is achieved by detecting anomalies in real-time, minimising the need for on-site human input. Therefore, this study conducts a long-term vibration monitoring of a three span historical masonry arch bridge by installing tri-axial accelerometers on each pier of the upstream spandrel wall. The modal parameters are identified using a Fast Bayesian Fast Fourier Transform method. This study shows the time-series of identified frequencies in the transverse direction of the right pier to discuss the daily trends of the identified frequencies and the identification uncertainty in normal period.