The introduction of the article presents the problem of interpreting the level of fire hazard basing on Graham’s ratio, which, in certain ranges of the value of its denominator, may be wrong. The range of credibility for the index is also discussed. The issue of nitrogen inertisation and its influence on the value of the discussed index is also addressed. To determine the influence, two statistical samples were set. They consisted of the results of precise chromatographic analyses of the air samples collected in the longwall areas which were not subjected to inertisation and in the areas where nitrogen was applied as the inert gas. Then, with Student’s t-test, there was conducted a comparative analysis of both groups with regard to the equality of the average concentrations of gases emitted in the coal self-heating process. At the end, there were developed criteria for the application of Graham’s ratio for the air samples of the increased content of nitrogen, which, according to the discussed index, did not indicate the occurrence of an endogenous fire hazard.