Because of continuous blood transfusions, thalassemia patients are subjected to peroxidative tissue injury by the secondary iron overload. In accordance, analysis of serum from 42 beta-thalassemia patients, aged 4 to 40 years, showed that the mean concentrations of conjugated diene lipid hydroperoxides (CD), lipoperoxides evaluated as malondialdehyde/ thiobarbituric acid (MDA/TBA) adducts, and protein carbonyls increased about twofold with respect to control. Ferritin levels were positively correlated with the amount of MDA (r = .41; P = .007) and showed a positive trend with CD (r = .31; P = .07) and protein carbonyls (r = .35; P = .054), as further evidence of the deleterious effects of high tissue iron levels. Marked changes in the antioxidant pattern were also observed in all patients. Evidence is presented of a net drop in the concentration of ascorbate (-44%), vitamin E (-42%), vitamin A(-44%), beta-carotene (-29%), and lycopene (-67%). On the other hand, an increase of uric acid and bilirubin was observed, whereas serum albumin and glutathione were in the normal range in all patients. As a result, the total serum antioxidant potential, measured as trolox equivalent antioxidant capacity appeared significantly decreased by 14%. Serum levels of vitamin E were inversely correlated with ferritin (r = -.45; P = .003), suggesting a major consumption of this antioxidant under iron overload. Nontransferrin bound iron (NTBI) was in the range 4.5 to 54.8 micrograms/dL (mean, 21.8 +/- 13.9). Although NTBI had a positive trend with ferritin (r = .37, P = .03), no clear correlation was found with either MDA or vitamin E. A mild to severe hepatic damage, as assessed by serum transaminases, was shown in 24 of 42 patients. Serum levels of vitamin E (r = -.49, P = .015), vitamin A (r = -.48, P = .016) and lycopene (r = -.47, P = .020), were inversely correlated with the levels of transminases. On the other hand, lipid-soluble antioxidants in thalassemia patients were depleted to the same extent in hepatitis C virus (HCV)-infected (31 subjects) and in HCV-uninfected (10 subjects), while in the normal range in serum from 30 nonthalassemic patients with HCV-related chronic hepatitis. These results point out that the iron-induced liver damage in thalassemia may play a major role in the depletion of lipid-soluble antioxidants. The variations of the parameters evaluated in the present study were not correlated with the age of the patients. Our results suggest that the measurement of peroxidation products, matched with evaluation of antioxidants, may be a simple measure of iron toxicity in thalessemia, in addition to the conventional indices of iron status.