Abstract: Hydraulic shearing is an important method in the development of hot dry rock geothermal energy. Studying the differences in the acoustic emission spectrum during fault slip in hot dry rock reservoirs with and without fluid involvement is of great significance for the impact range of hydraulic shearing engineering in hot dry rock geothermal exploitation and the site selection of production wells. To gain a deeper understanding of the acoustic emission frequency spectral characteristics during the shear slip process of granite fractures under different hydraulic coupling effects, a self-developed high-confining pressure large-displacement real-time dynamic shear-seepage testing device is used to conduct shear tests on dry and water-injected granite fracture samples, with the water in jection pressure gradient set to 5, 6, and 8 MPa. The results indicate that normalized main frequency amplitude and ring count rate can effectively characterize the sliding characteristics during water injection. The acoustic emission signals produced by the dry fracture surfaces during the stick-slip process have significantly higher normalized main frequency amplitude and ring count rate compared to those of water-injected fracture surfaces. Moreover, as the water injection pressure increases, these two parameters show a negative correlation with the water injection pressure. Spectral analysis results indicate that water injection reduces the main frequency and amplitude of the acoustic emission signals. The acoustic emission signals of water-lubricated friction are generally lower in these two parameters compared to dry friction, and both the main frequency and amplitude decrease as the water injection pressure increases. The frequency of low-amplitude acoustic emission signals and the main frequency are both affected by the water injection pressure, with the frequency of acoustic emission signals increasing as the water injection pressure increases, while the main frequency shows a decreasing trend. The reduction in the intensity of AE signals and the increase in the number of low-amplitude AE signals can be used as criteria to determine whether there is fluid involvement in fault slip.