Abstract: Coal mine safety monitoring system is the core link to ensure production of coal mine safety. Based on the functional safety theory, combined with the International Standard IEC 61508 and the coal mine industry specification, this paper systematically studies the functional safety requirements, classification methods and evaluation process of the safety monitoring system. Firstly, the functional safety requirements are put forward from four aspects of real-time monitoring, redundancy design, fault diagnosis and emergency response, and the monitoring threshold and fault tolerance mechanism of key parameters such as gas concentration, roof pressure and ventilation state are defined. Secondly, combined with the risk matrix method and the layer of protection analysis （LOPA）, a safety integrity level （SIL） classification model with risk occurrence probability （P） and consequence severity （S） as dimensions is established, and the failure probability of the system is quantified by fault tree analysis （FTA）. Finally, taking the gas monitoring system of a coal mine as an example, the applicability of the classification method is verified, and suggestions on sensor redundancy configuration and algorithm optimization are put forward. The research shows that the functional safety classification can effectively improve the reliability and risk control ability of the monitoring system, and provide support for the intelligent transformation of coal mines.