Abstract: Risetime/amplitude (RA) and average frequency (AF) have been usually used for qualitative analysis of fracture mechanism in acoustic emission (AE) monitoring. However, regardless whether fracture is shear or tensile in macroscopic view, it can be observed in laboratory experiments that the AE signals of shear increases when it is close to the failure stage of specimens. Therefore, RA and AF may also have the potential in indicating the violent reputure of rock. Furthermore, as the value of RA would increase with the distance within some limits, the observed RA and AF would be closer to the shear feature, which means the index is relatively safer under attenuation and is appropriate for in-situ monitoring. Based on the data monitoring of Huayingshan Tunnel, Yuguang Expressway during the construction process, the distributions of RA and AF on positions of different distances of the seismic source were compared. Results show that the maximum value of RA increases distinctly with the distance increase between the sensors and the seismic source, whereas the distribution of values of AF are nearly the same at different distances. To verify the validity of RA and AF in indicating the rupture of rock, parameter r of RA/AF ratio was set and the time history of r and coefficient of variation (CV) of r during the rupture process were studied and compared with other regular indexes, such as absolute energy and b value. The variation of CV could describe the intense rupture of rock properly and the analysis of CV could get a safer evaluation result especially when dealing with small-scale failure in rock mass. To find the best statistical method of CV, three statistical methods of CV were compared and results show that the CV of r can well illustrate the rock rupture, CV₁ is more suitable for situations that the AE signals vary and are discrete, and CV₃ is appropriate for monitoring of continuous AE signals.