Acoustic Emission (AE) analysis has been established as a valuable monitoring technique for detecting wire breaks, for example, in assessing the risk of Stress Corrosion Cracking (SCC) in older prestressed concrete bridges. A broad knowledge of the characteristics of the damage signal is crucial to reliable detection and subsequent classification. For this reason, extensive experiments were carried out on six prestressed concrete girders from three different bridges susceptible to SCC. Of these girders over 382 wire breaks were provoked under a variety of boundary conditions, and a database of 2700 signals was collected. Typical AE features were calculated from the signals and used for analysis. For classification purposes reference data from real traffic noise and rebound hammer impacts were added to the database. The distributions of wire break features were found to be generally similar for all girders. Further, the local bonding conditions between the wire and the grout at the break location showed to affect the signals respective energy, which was mainly released below 50 kHz. The best separation of damage and noise signals was found for amplitude, energy, and RMS. Statistical parameters were extracted for use in deriving valid thresholds for monitoring applications.