Abstract: Coal and gas outburst is a major geological hazard during mine excavation, and the effectiveness of its prevention and control largely depends on the accurate prediction of gas-enriched zones. Geophysical exploration plays a crucial role in gas outburst control due to its non-invasive nature and high resolution. This paper begins with an overview of current prediction methods for gas-enriched zones, followed by a systematic review of the development of geophysical detection technologies in this field. In the early stages, surface 3D seismic exploration was predominantly employed, which initially enabled the prediction of the spatial distribution of gas-enriched zones by establishing high-density observation systems and integrating multi-dimensional data processing methods. However, as mining depths continue to increase, surface exploration techniques have become inadequate for deep mining due to declining detection accuracy. This limitation prompted the development of geophysical detection methods within mine roadways, giving rise to techniques such as in-seam seismic wave exploration and radio wave penetration. By relocating detection sources to near-field positions underground, these methods significantly improved detection accuracy. Nevertheless, metal interference and high background noise in the mine environment substantially increased the complexity of data interpretation. Against this backdrop, research focus has gradually shifted to borehole geophysical detection methods in mines. Through near-field signal acquisition, external interference is effectively suppressed. The introduction of borehole geophysical methods such as nuclear magnetic resonance logging and borehole radar imaging allows for detailed characterization of coal seam fracture development and its dynamic correlation with gas adsorption–desorption processes, providing reliable technical support for the precise identification and evaluation of gas-enriched zones. Future research will aim to establish a technical framework characterized by “full-space detection and multi-method collaboration,” with an emphasis on developing risk early-warning models based on 3D dynamic evolution to enable intelligent identification of coal and gas outburst hazardous areas. By integrating multi-source geophysical field data fusion and deep learning technologies, high-precision identification methods for geological anomalies will be developed. This will continuously enhance the spatial resolution of mine dynamic disaster prevention and control to millimeter-level standards, while also establishing a real-time monitoring system and intelligent decision-making platform for the safe exploitation of deep coal resources, thereby providing full lifecycle engineering support for mining process optimization.