In the present work the application of acoustic emission technique has been proposed as a tool for the evaluation of corrosion processes occurring under SCC conditions on prestressing steel strands. To simulate the possible corrosive environment under service conditions experimental tests were performed with different electrolytes and electrochemical conditions (open circuit potential, cathodic and anodic polarisation). The use of multi-variable techniques, coupled to a classic statistical approach, has permitted to discriminate separated clusters in the population of the acquired AE signals related with specific damage events. Thereby the proposed approach allowed to create corrosion maps able to classify both global and punctual degradation phenomena. In particular different areas of the maps were related to the different corrosion mechanisms experimentally observed during the SCC tests.