In the lowermost rocks of the Rustenburg Layered Suite (Bushveld Complex), levels of total Cr2O3 are up to ten times higher than its putative solubility (0.16% Cr2O3) in the magmas that many investigators have accepted as parental to the dunites, harzburgites, pyroxenites, and norites of the Lower, Lower Critical, and Upper Critical zones. Various existing models fail to explain this “chromium paradox” in so far as they may explain why chromite is precipitated but do not account for its abundance in many cumulates. Simple mixing calculations and quantitative Adiabat_1ph runs support our model in which a heterogeneous (liquid + crystals) system was evolved in a staging chamber beneath the complex by mixing of komatiitic melts with 20 to 30% of crustal melts at depths appropriate to pressures between 10 and 4.5 kbars. We contend that crystal settling and the effects of elutriation during magma ascent had generated a high SiO2 cap with modal olivine, orthopyroxene, and chromite increasing with depth in the staging chamber. The magma from the high SiO2 cap of the staging chamber was first injected into the largely sedimentary Transvaal Supergroup, generating what is now termed the Marginal zone. Slurries of crystals plus liquid from the spasmodically rejuvenated main body of the staging chamber were then injected into the shallower Bushveld chamber to form the Lower and Critical zones. Chromite within the slurries introduced in this way could account for their high gross Cr content. Frequent reversals of normal evolutionary trends evident in the cumulates are attributable to the episodic nature of the influxes from the staging chamber. The lithologically diverse nature of the floors upon which the chromitite layers accumulated (peridotites, pyroxenites, norites, and anorthosites) was therefore more a function of the timing of emplacement of the Cr-rich slurries generated in the deeper reservoir than of specific evolutionary processes within the Bushveld chamber itself.