Phenolic compounds in red wine can exert antioxidant effects on in vitro lipoprotein oxidation. This has led to speculation that red wine consumption mediates unique anti-atherosclerotic effects compared to other alcoholic beverages. However, studies assessing the effects of red wine consumption on lipoprotein oxidation ex vivo have not been conclusive. The recent identification of the F2-isoprostanes as oxidative products of arachidonic acid has provided a reliable measure of in vivo lipid peroxidation. This randomized trial investigated changes in plasma and urinary F2-isoprostane concentrations following red wine, white wine, or dealcoholized red wine consumption in humans. Eighteen male smokers consumed, in random order, red wine, white wine, or dealcoholized red wine, for two weeks with one week washout between beverages. Plasma and urinary F2-isoprostane concentrations were measured before and after each beverage. Serum gamma-glutamyl transpeptidase (gamma-GT) and urinary 4-O -methylgallic acid were measured as markers of alcohol consumption and phenolic acid absorption, respectively. Plasma F2-isoprostanes (p < .05) decreased significantly with dealcoholized red wine but not with the alcohol-containing beverages. Urinary excretion of F2-isoprostanes showed a similar trend. gamma-GT decreased significantly with dealcoholized red wine and increased with both alcohol-containing beverages (p < .01). Urinary excretion of 4-O-methylgallic acid increased significantly (p < .001) in the 24 h urine samples following red wine or dealcoholized red wine ingestion, but not with white wine. Serum urate increased and beta-carotene decreased with both alcoholic beverages relative to dealcoholized red wine. There was no change in the antioxidants alpha- and gamma-tocopherol or vitamin C with any of the beverages. The results suggest that polyphenols in dealcoholized red wine can reduce in vivo lipid peroxidation as measured by F2-isoprostanes in smoking subjects. However, no reduction in lipid peroxidation was observed following red or white wine consumption, suggesting that any protective effects of wine drinking on cardiovascular disease are unlikely to be related to inhibition of lipid oxidation.