With the availability of iron as iron-dextran, possessing great stability and correspondingly low toxicity, it has become possible to achieve in animals, within a short space of time, levels of iron overload greatly in excess of those attained in human siderotic states. The quantitative aspects of iron loading of this kind have been studied in four species of normal and iron-laden animal. In the rat a high proportion of the iron remains at the injection site whereas in mice, rabbits, and dogs more is withdrawn to the liver. Biochemical changes in the liver and at the injection site include raised levels of acid phosphatase and of material reacting with thiobarbituric acid, probably consisting of lipoperoxides. Elevated serum transaminase levels were also observed, even when iron-dextran was administered by the intravenous route. These biochemical effects reflect not only the systemic swamping with iron, but also the profound local changes at the site of injection, consisting of foci of chronic inflammation, muscle necrosis, fibrosis, and massive deposits of iron-laden macrophages that completely transform the injected tissues. A number of long-term experiments are reported in which the object has been to induce sarcomas in iron-laden rats and mice. The relationship between the dosage of iron-dextran used and the yield of sarcomas at the site of injection makes it likely that a threshold dose exists below which injection-site sarcomas do not arise because the dose is too low to elicit either the characteristic local changes in the injected tissues or the systemic effects. Dosage must be considered in relation to the available mass of tissue at the injection site. The development of neoplasia may be connected in part with the prolonged reticuloendothelial stimulation induced by large, but not by moderate, doses of iron-dextran.