This paper presents a statistical relationship between the pyrite particle size distribution and the potential amount of pyritic sulfur reduction achieved by specific-gravity-based separation. This relationship is obtained from data on 26 Ohio coal samples crushed to 14 ?? 28 mesh. In this paper a prediction equation is developed that considers the complete statistical distribution of all the pyrite particle sizes in the coal sample. Assuming that pyrite particles occurring in coal have a lognormal distribution, the information about the particle size distribution can be encapsulated in terms of two parameters only, the mean and the standard deviation of the logarithms of the grain diameters. When the pyritic sulfur reductions of the 26 coal samples are related to these two parameters, a very satisfactory regression equation (R2 = 0.91) results. This equation shows that information on both these parameters is needed for an accurate prediction of potential sulfur reduction, and that the mean and the standard deviation interact negatively insofar as their influence on pyritic sulfur reduction is concerned. ?? 1988.