A theoretical analysis was performed to determine the number of fractions a multidisperse, immunomagnetically labeled cell population can be separated into based on the surface marker (antigen) density. A number of assumptions were made in this analysis : that there is a proportionality between the number of surface markers on the cell surface and the number of immunomagnetic labels bound; that this surface marker density is independent of the cell diameter; and that there is only the presence of magnetic and drag forces acting on the cell. Due to the normal distribution of cell diameters, a "randomizing" effect enters into the analysis, and an analogy between the "theoretical plate" analysis of distillation, adsorption, and chromatography can be made. Using the experimentally determined, normal distribution of cell diameters for human lymphocytes and a breast cancer cell line, and fluorescent activated cell screening data of specific surface marker distributions, examples of theoretical plate calculations were made and discussed.