In order to evaluate the exploitation of broad-band seismic networks for landslide characterization, we selected a number of well known mass movements occurred in the Italian Alps, characterized by volumes ranging between 25.000 and 34.000.000 cubic meters and different failure dynamics. We analyzed the related seismic signals recorded by seismic stations located, on average, 50 km away from the sources, focusing the study on the 1–30 Hz frequency band. The time frequency domain analysis of data allowed identifying some common signal characteristics, including emergent onsets on all channels, slowly decaying tails and a triangular spectrogram shape. Moreover, the time series of seismic record allowed identifying the different sub-events in which it is possible to divide a landslide, giving a first qualitative characterization of these phenomena from seismic records. In this work, the selected events are described and the associated seismograms are presented and analyzed together with a first discussion of their spectral characteristics.