It is known (Eckhoff, 2003) that an experimental study of aero-suspension of dust with low explosivity in a 20-liter chamber leads to overestimation of the explosion. A special concern is the risk of a qualitative error, when non-explosive dust will be transferred to explosive dusts, which will cause unjustified costs for ensuring the safety of industries involving this dust. This work is aimed at reducing this risk. In this work, a quantitative analysis of the causes of overstatement of dust explosiveness is performed. It is shown that the testing of dust/air mixture in a 20-liter chamber does not correspond to the normal initial conditions of the investigation (pressure 100 kPa, temperature 25 °C) stated in the methodologies and, in fact, is an explosion hazard study of dust/air mixture with an increased initial temperature and an increased initial pressure in the chamber. Two processes lead to an increase in the initial temperature: the dispersion of particulate material in the chamber by a pulse of compressed air from the receiver; adiabatic compression of dust/air mixture upon activation of the ignition source and local burning out of the dust found in the flame and/or near the flame of the ignition source. The latter process leads to an increase in the initial pressure in the chamber. The implementation of this analysis required the development of a reliable criterion for the explosion of dust, since there is still no single idea of such a criterion, judging by the norms of the United States and European countries. The new criterion is based on two assumptions: (1) on limiting the variety of the development of dust ignition in two scenarios (Cashdollar and Chatrathi, 1993) - local burning out of dust in some neighborhood of the ignition source and dust explosion, covering the entire volume of the chamber and (2) on the essential difference between pressure jumps in the chamber, expected for different scenarios. Two variants of application of the results of this work are demonstrated. First, it is possible to forecast the conditions under which an explosive danger arises in the dust, which is not explosive under normal conditions. Such a forecast is made in case of recording the explosion of this dust in a 20-liter chamber and assessing the real initial conditions of the study. In particular, an explosion of anthracite, investigated in a 20-liter US Bureau of Mines, is predicted to be explosive at a temperature of 140 °C. Secondly, it is possible to outline ways of realizing the conditions for experimental investigation of dust with a low explosivity, close to normal. They will significantly improve the reliability of the conclusion about the low explosiveness of combustible dust without the use of large-scale equipment. Two such methods are proposed in the work. The first method is based on the reasonable assumption that for a dust with a low explosion hazard low oxygen concentration LOC » 0.21, and the known empirical linear dependence of the explosion index Kst on the oxygen content in air. Within the framework of this method, a search is made for the LOC, by examining Kst for dust suspensions in air enriched with oxygen. The explosion hazard of dust is judged by the ratio between the extrapolation obtained by the LOC and the usual oxygen content in the air (0.21). The second method involves a modified design of a 20-l camera, which differs from the standard design of a 20-liter chamber in a vertically extended shape and variable volume. The latter is achieved by using a “flexible” top end of a polymer film, initially concave into the chamber, but assuming a convex shape after dust dispersion and triggering the ignition source.