It is well established that vegetables and fruit (VF) contain antioxidant phytochemicals. Consequently, it is expected that individuals who consume diets with a high content of VF should be better protected against oxidative cellular damage than individuals who do not, but not all data support this assumption. The objective of this study was to identify possible explanations for this conundrum. The effects of two diets that differed in VF content on markers of oxidative damage were studied. Sixty-four women participated in a 14-day dietary intervention. Subjects consumed on average either 3.6 or 12.1 servings of VF per day. The primary end points assessed were 8-hydroxy-2-deoxyguanosine (8-oxo-dG) in peripheral lymphocyte DNA and 8-isoprostaglandin F-2alpha (8-iso-PGF2alpha) excreted in urine. Subjects consuming the high versus low VF diet had lower concentrations of 8-oxo-dG (p < 0.01) and of 8-iso-PGF2alpha (p < 0.01). However, the reduction in oxidative end points by high VF was not uniform. Rather, an antioxidant effect was observed primarily in individuals whose oxidative end points at baseline were above the median for the study population. Using change in plasma carotenoids (end point minus baseline measurements) as an index of phytochemical intake, the reduction in oxidative markers was inversely proportional to change in plasma carotenoids; this effect was stronger for lipid peroxidation (p < 0.01) than DNA oxidation (p < 0.05). These findings imply that increasing exogenous antioxidant exposure may primarily benefit individuals with elevated levels of oxidative stress. Null findings do not necessarily indicate that an antioxidant compound lacks in vivo activity.