Monitoring represents one solution for the safeguard of historical buildings. The need for a non-destructive and comprehensive monitoring methodology suggests using related to Structural Health Monitoring. This paper is intended to present the outcomes of an experimental campaign on a masonry triumphal arch representing a real scale model of a church part, which was built outside ELSA laboratory at the Joint Research Centre of European Commission. This study aims to evaluate the damage pattern of the structure through simplified dynamic methods producing a quick evaluation of structural safety, easy to use on real cases. As in traditional monitoring, both the instrumentation precision and the measurement variability due to the different testing condition (e.g. ambient conditions) have to be considered. The related effects on the structural dynamic behaviour were analysed and evaluated in order to distinguish an effective change in the “structural health” (a real damage) from an alteration caused by external conditions (a “false positive”). Once studied such effects, settlements were induced to one column base through an “ad hoc” device. Varying the settlement width, three damage levels were obtained in the structure. For each state the structural dynamic properties and their variation were evaluated. Sensitivity of dynamic behaviour to structural damage and to its changes was analysed comparing the results for each level.