Oxidative stress has been suggested as one of the physiopathologic conditions underlying the association of total plasma homocysteine (p-tHcy) with cardiovascular disease (CVD), but this hypothesis has not been validated in human epidemiological studies. We measured plasma and erythrocyte antioxidant enzymes glutathione peroxidase (GPx) and superoxide dismutase (SOD), along with serum lipid-soluble antioxidants alpha-tocopherol, beta-carotene, lycopene and retinol, in a sample of 123 healthy elderly subjects (54 men, 69 women). Plasma malondialdehyde (p-MDA) was determined as a marker of lipid peroxidation, and p-tHcy was quantified by HPLC. No significant differences were found for p-MDA, GPx or SOD activities or serum antioxidant concentrations, in subjects with elevated p-tHcy (> or =15 micromol/l) as compared to those with lower plasma homocysteine. Hyperhomocysteinemia did not lead to increased risk of having the highest p-MDA values, in either sex. We found no evidence that p-tHcy was associated with lipid peroxidation in this elderly human sample. Our results do not support the view that hyperhomocysteinemia would induce an adaptive response of antioxidant systems, either. More epidemiologic and clinical research is needed to clarify whether homocysteine promotes atherosclerosis by means of an oxidative stress mechanism.