A recently improved theory for the distribution of resolution in complex separations described by statistical-overlap theory is used to predict the average numbers of singlet, doublet, and triplet peaks in such separations, even when the peak capacity is much less than the number of components requiring separation. The theory is fundamental and modifies previously published equations for these numbers by causing the average minimum resolution that defines saturation to depend on saturation itself. Interestingly, theory shows that the resolutions describing the separation and overlap of single-component peaks in singlets, doublets, and triplets actually differ. The theory predicts correctly the numbers of singlets, doublets, and triplets in computer simulations of separations, even at high saturation. Example calculations are provided to show the ease with which the theory can be used. (C) 1999 Elsevier Science B.V. All rights reserved.