To elucidate the effects of iron on lipid peroxidation, three kinds of assays were done. (i) The effects of intravenous injection of single doses of iron dextran (100 mg iron) on lipid peroxidation in various tissues and the blood plasma of rats were examined by the thiobarbituric acid reaction. Malondialdehyde concentrations were significantly elevated in the spleen, heart, and plasma 3 h after injection, whereas significant increases were observed in the liver and adipose tissue at 24 and 48 h, respectively. In the liver and spleen, the elevated malondialdehyde concentrations persisted until Day 28. In contrast, levels were reduced in the heart and adipose tissue within 4 weeks after iron injection. Plasma malondialdehyde concentrations were 70-times that of controls at 24 h after iron injection. The level subsequently decreased sharply by Day 6. In red blood cells, lipid peroxidation was not affected by iron. Malondialdehyde levels were correlated with the iron contents of the liver, spleen, heart, adipose tissue, and plasma (r value range 0.39-0.88, p less than 0.05). Furthermore, there was a strong correlation in the liver and spleen at iron levels below 2,000 micrograms/g (r = 0.94, p less than 0.0001; r = 0.94, p less than 0.0001, respectively). (ii) In vitro experiments demonstrated that the addition of iron as ferric chloride, iron dextran, ferritin, and hemosiderin to normal liver homogenate accelerated malondialdehyde production. However, such increases were less than 10% of those caused by equivalent iron in the liver homogenate of iron treated rats. (iii) Compared to controls, spleens from eight thalassemia patients showed a high level of malondialdehyde and iron.