Evaluation of self-ignition hazard of bulk materials requires experimental determination of self-ignition temperatures as a function of volume. There are two standardised methods: (1) determination of the self-ignition temperature of dust samples in oven and (2) measurement of the self-ignition temperature of a dust layer deposited on a hot surface. Sometimes, the sample behaviour during these tests makes the second method difficult to apply. The self-ignition phenomena in these two tests rely on the same principles. Their results are interpreted with the help of theoretical relations. The correlation described in this paper can be considered acceptable to deduce self-ignition temperature of a dust layer, based on results of self-ignition of the same dust in heating ovens, if the Biot number (alpha) can be estimated. Uncertainty on the correlation is near 30 K. This uncertainty is on the same order of magnitude as the difference in the self-ignition temperature on a hot surface for thickness between 5 and 15 mm. (C) 2008 Elsevier B.V. All rights reserved.