Abstract: This study focuses on the Upper Zone of the Čukaru Peki Copper-Gold Mine in Serbia and addresses two key issues: how to achieve effective coverage and reliable monitoring of critical zones using a microseismic monitoring network, and how to identify precursory signals of rock-mass instability based on microseismic parameters. A 32-channel online microseismic monitoring system was established, and the sensor layout, source-location accuracy, and magnitude detection capability for the −200 m to −70 m levels were evaluated through numerical simulation. Based on field monitoring data, the spatiotemporal evolution of microseismic events under mining disturbance was analyzed, and a joint identification method using the energy index （EI） and cumulative apparent volume （CAV） was employed. The results show that the optimized monitoring network can effectively cover the UZ main orebody and the UZ−2 zone, with planar location accuracy better than 15 m in the key monitoring area and a minimum detectable magnitude of approximately −2.3. Temporally, microseismic activity is strongly controlled by blasting operations and exhibits a daily distribution pattern of mainshocks and aftershocks. Spatially, events are mainly concentrated in active stoping areas, fault intersection zones, and lithological transition zones. From December 25 to December 31, 2023, EI shows a continuous decline whereas CAV increases rapidly, indicating that the surrounding rock mass transitions from a stable energy accumulation stage into an unstable damage evolution stage. Field verification shows that the combined EI-CAV indicator can provide at least 1 d of advance warning of relatively large-scale rock-mass fracturing. The study indicates that microseismic monitoring combined with EI-CAV joint identification can effectively characterize ground pressure activity patterns under mining disturbance and provide a technical basis for ground pressure early warning in deep underground mines.