Accurate source event location is important in fracturing monitoring and characterization. Velocity anisotropy has a great influence on both events matching and events location. Failure to take into account the velocity anisotropy can lead to huge errors in locating events. In this article, we have presented an experimental study on lower Silurian shale from the Sichuan Basin. The experimental observations include ultrasonic measurements, acoustic emissions (AEs) in a three-point bend experiment, and CT scanning of the original sample and the fractured sample. The ultrasonic measurements show that the shale sample has strong velocity anisotropy. Initially, AEs are analyzed using the conventional event-matching method and event location method (Geiger’s method), and the detected events are compared to the X-ray image of the fracture. Event-matching aims to obtain AE signals from the same source event and thus assists in selecting valid AE signals that come from the same source and are received by at least four sensors, to determine the location of the source. Although many reliable signals are obtained by isotropic event-matching, fewer sources were located than expected, and the event location results did not match the fracture distribution. To address this problem, an improved event-matching method is proposed using a stricter matching threshold based on directional velocity rather than a single threshold same for all directions. In addition, we propose an improved Geiger’s method using the anisotropic velocity model. The new methods located more sources that better match fracture distribution than the results of the isotropic method. We have concluded that both event-matching and the source location of the fracturing are largely influenced by velocity anisotropy, and thus in practice, the velocity anisotropy information obtained from various measurements (e.g., laboratory measurements, well logs, VSP, and velocity analysis of reflected seismic surveys) should be involved in both processing procedures. This study can be useful to provide some background for monitoring and predicting dynamic geo-hazards in relation to the AE method.