Analytical expressions for the floor spectra evaluation play a key role for a correct definition of the seismic input induced to non-structural elements or local mechanisms in existing buildings. They have to be able to properly assess the possible amplification phenomena, but also to correctly describe the effects of nonlinearities due to structural damage. Due to the complexity of such phenomena, data on existing structures hit by earthquakes constitute a precious source for a better understanding of the topic and the validation of analytical expressions. In this framework, the paper aim is twofold. On one hand, it evaluates the entity of seismic amplification through experimental evidence from in-situ measurements on existing monitored structures. On the other hand, it presents the application of an analytical expression for the floor spectra already developed by the Authors to two unreinforced masonry (URM) buildings. The case-studies are the former Fabriano courthouse (Ancona, Italy) and the Pizzoli’s town hall (L’Aquila, Italy). They were both hit by the 2016/2017 earthquake in Central Italy and are permanently monitored by the Italian seismic monitoring system of the Italian Department of Civil Protection (DPC). With the aim of the validation, the paper shows the comparison between experimental and analytical floor spectra for various minor events and mainshocks of the Central Italy earthquake. Since the two case-studies exhibited different damage level (from slight to moderate, respectively), the comparison allowed to verify the reliability of the expression both in the pseudo-elastic and moderate nonlinear fields.