As a key component of friction stir joining (FSJ), tools play an important role in welding quality and welding performance. In this study, the acoustic emission (AE) signal acquisition system was established. The AE signals from FSJ of SiCp/Al composites were analyzed in the time domain, frequency domain, and time-frequency domain respectively. The root mean square (RMS) value, spectrum distribution, and energy distribution of AE signals during FSJ were obtained. The relationship between the tool wear state and the AE signal was studied. The results show that with the increase of welding times, the wear degree of the pin is aggravating. The more serious the tool was worn, the greater the RMS value of the acoustic emission signal was. With the increase of wear, the energy of AE signals gradually moves to the low-frequency region, and the spectrum tends to be concentrated. According to wavelet packet decomposition of AE signals, it has been found that in the frequency bands 3-8, the energy percentage increases with the aggravation of agitation tool wear, while the opposite trend appears in the frequency bands 10-16. [ABSTRACT FROM AUTHOR]