Abstract: In order to avoid the hazards to steel components caused by welding slag defects, acoustic emisson (AE) signals were colleted during the deformation and fracture process of the welding slag specimen and analysis was used to analyze the AE signal and to extract the distribution range of the characteristic parameters. By comparing the characteristic parameters of the standard specimen and welding slag specimen, it was found that the characteristic parameters like local spectrum 1 and 3, frequency centroid, average frequency and peak frequency could be used to distinguish standard specimen and welding slag specimen. Present findings provide the judgment for the use of acoustic emission technology to detect the weld slag.