Abstract <p>The article presents a methodology for analyzing the reliability of geophysical methods that use portable instruments for forecasting dynamic phenomena. The need for developing this methodology is explained by the lack of recommendations in regulatory documents regarding the selection of the most appropriate method to forecast dynamic phenomena for specific conditions. This selection should be carried out by the relevant departments of coal-mining companies independently or with the help of scientific teams. Two portable devices for local prediction of rock bumps were chosen for the research: the Angel-M and Ripas devices, which implement the electromagnetic pulse method and a specific modification of the method based on parameters of artificial acoustic signals, respectively. The more reliable instrumental method based on the drilling fines yield was used as the reference one. It was assumed that the hazard level of the rock mass is the same when making predictions using the geophysical and instrumental methods. It can be quantitatively estimated as a ratio of the current value of the measured parameter to its threshold value. The reliability of the predictions made with portable devices was assessed by the correspondence of this ratio to the similar one calculated for the instrumental method. Testing of the proposed methodology at one of the Kuzbass coal mines showed insufficient reliability of predictions made using portable geophysical devices. This is probably explained by the following reasons: each of the devices does not control all of the key hazard factors of dynamic phenomena; there is no justification of the relationship between the hazard of the dynamic phenomena and the parameters controlled by the devices; when using the Ripas device, the geophone is pressed against the wall of the mine workings, so the fracturing within the near-wall rock mass significantly distorts the spectrum of the sounding acoustic signal.</p>