Supplemental beta-carotene has been shown to increase lung cancer risk in recent chemoprevention trials, especially in current smokers. Several possible mechanisms for this effect have been suggested based upon in vitro and animal studies, but mechanistic data from human studies to explain the excess risk are lacking. beta-Carotene has both antioxidant and prooxidant effects in vitro; therefore, we evaluated whether or not high-dose supplemental beta-carotene might have prooxidant effects in vivo, especially in current smokers taking high-dose supplemental beta-carotene for several years (median 4.0 yr). Urine samples (n = 55 total) were collected from both smokers and nonsmokers participating in a multiyear randomized chemoprevention trial of supplemental beta-carotene (50 mg/day) versus placebo. Samples were analyzed by GC/MS for total isoprostanes and for 8-iso-prostaglandin F2 (8-iso-PGF2), stable end products of lipid peroxidation in vivo. Smokers had higher levels of both total isoprostanes and 8-iso-PGF2. Smokers and nonsmokers randomized to beta-carotene had nonsignificantly lower concentrations of total isoprostanes and of 8-iso-PGF2 [mean +- SD 8-iso-PGF2/ml = 2.00 +- 1.72 (placebo smoker); 1.72 +- 1.66 (beta-carotene smoker); 1.22 +- 0.68 (placebo nonsmoker); 0.97 +- 0.62 (beta-carotene nonsmoker)]. These results indicate that supplemental beta-carotene, even when given at high doses for many years, does not have prooxidant effects in either smokers or nonsmokers, as measured by urinary excretion of F2-isoprostanes.