BACKGROUND: Antioxidant supplementation with vitamin E had no effect in the prevention of cardiovascular diseases (CVD) in three recent large, randomized clinical trials. In order to reassess critically the role of vitamin E in CVD prevention, it is important to establish whether these results are related to a lack of antioxidant action. METHODS: We examined the in vivo antioxidant effect of vitamin E (300 mg/day for about three years) in 144 participants in the Primary Prevention Project (females and males, aged >/= 50 y, with at least one major CV risk factor, but no history of CVD). Urinary 8-epi-PGF2alpha (isoprostane F2alpha-III or 15-F2t-isoP), a validated biomarker of lipid peroxidation, was measured by mass spectrometry. RESULTS: Urinary excretion of 8-epi-PGF2alpha [pg/mg creatinine, median (range)] was 141 (67-498) in treated and 148 (76-561) in untreated subjects (p = 0.10). Taking into account possible confounding variables, multiple regression analysis confirmed that vitamin E had no significant effect on this biomarker. Levels of 8-epi-PGF2alpha were in the normal range for most subjects, except smokers and those with uncontrolled blood pressure or hyperglycemia. CONCLUSIONS: Prolonged vitamin E supplementation did not reduce lipid peroxidation in subjects with major cardiovascular risk factors. The observation that the rate of lipid peroxidation was near normal in a large proportion of subjects may help explain why vitamin E was not effective as an antioxidant in the PPP study and was ineffective for CVD prevention in large scale trials.