To prevent emergencies at mining enterprises, a model of controlling stability of geotechnical system by fuzzy logic methods was developed based on the data fuzzification, inference and defuzzification. The main principles were formulated for the identification of specific features of dangerous production facilities in mining industry. To determine the input parameters for the fuzzy logic model, the systematization of uncontrolled (mining and geological) and controlled (technical and technological) parameters, the presence or change of which affects the stability of the rock massif and roadways, was carried out. The application of such methods as aggregation, implication, defuzzification was substantiated for processing input signals. An algorithm of fuzzy logic inference was formed to control parameters of a geotechnical system. The model differs by its ability to adapt to the specifics of controlling the geotechnical system “support - rock massif” and to select logical rules depending on the established criteria.