ABSTRACTThe parameters which characterize an explosive are derived from laboratory tests and none is directly usable for seismic purposes. Another one, derived from field experiments is proposed in this paper. It allows the comparison of two different explosives. The test must be conducted with care. Precautions concerning the equipment and the shot point detailed. The first available results suggest that the new parameter does not depend on the size of the charge nor on the shooting conditions, though the experimental data are not numerous enough to allow a general statement. It is shown how a coefficient of seismic efficiency can be used for selecting an explosive; and considerable amounts of money can be saved by the best choice. More studies on the generation of artificial seismic waves would be highly profitable.