<p>Prediction of coal bump is of paramount importance to ensure the safety of underground workplaces, especially for deep-seated coal seams. The occurrence of coal bumps is influenced by various factors, including geological, geotechnical, and operational parameters. Most of the existing indices are developed for intact rock samples at laboratory scale study, which does not represent the rock mass conditions at the field. Moreover, these indices do not consider the effects of different mining operational parameters on the coal bump. In this study, an index of burst energy coefficient has been developed to predict coal bump by considering the intrinsic parameters as well as operation parameters. Based on the developed index, an empirical model has been developed to classify the coal bump as expressed by the probability of its occurrence. A parametric study by numerical modelling is carried out to obtain the stress–strain curve under uniaxial compressive strength testing of coal pillars. It is found that the depth of cover is the most influential parameter of burst energy coefficient vis-à-vis coal bump. Other significant parameters are RMR, w/h ratio of coal pillars and Young’s modulus of roof and coal. The study shows the increase in depth of cover and the presence of strong competent roof strata increase the chances of coal bump. The results of the parametric study are used to develop predictive models for the burst energy coefficient and the probability of coal bump occurrence. The predictive models are validated by the coal bump events at different underground coal mines and by comparing them with an existing index. The uniqueness of the models is their applicability in field conditions rather than laboratory conditions and to classify the coal bump events in terms of probability of occurrence. The predictive models would be easy-to-use tools for coal bump in underground coal mines.</p>