Previously reported animal models of hemosiderosis fall short of simulating the human disease state of transfusion-induced hemosiderosis. An explanation for this was found by reexamining heat-treated red blood cell (rbc) loading in the mouse. After 18 intraperitoneal transfusions, each equal to two-thirds of mouse rbc volume, liver Fe reaches a level of 0.3% Fe (dry weight), which is far below the 2%-8% found in heavily transfused patients. Using the easily synthesized chelate compound ferric acetohydroxamate, Fe(AHA)3, liver Fe levels in the rat of up to 2% were achieved after 38 intraperitoneal injections. Fe was distributed in both the reticuloendothelial (RE) system and parenchymal cells, as ascertained by light microscopy. No definite histological or biochemical abnormalities could be demonstrated in loaded rats. Cardiac Fe was approximately doubled. Thus, chelate loading, while producing Fe liver levels similar to those of humans with hemosiderosis, may still be of limited usefulness in studying long-term sequelae. On the other hand, this model can be used in determining responses to chelating agents. To explore this, Fe stores were first labeled by giving 59Fe as 59Fe(AHA)3 prior to loading. In animals loaded to 0.7% liver Fe (calculated), desferroxamine, at a dose of 400 mg/kg, induced a 20-fold rise in urinary Fe. This was duplicated by AHA at a dose of 800-1600 mg/kg. It is concluded that Fe(AHA)3-loaded rats are a potentially useful model of hemosiderosis and that further studies are needed to determine whether AHA can be effective in the treatment of transfusion-induced hemosiderosis.