This study evaluated how prostatic levels of antioxidants relate to plasma levels and self-reported usual dietary intake. Definition of these relations may aid in interpreting studies of antioxidant exposure and prostate cancer risk. Between July 1996 and April 1997, plasma and prostatic tissue levels of tocopherols, carotenoids, and retinol were measured in 47 men undergoing radical prostatectomy or transurethral prostatectomy at Loyola University Medical Center, Maywood, Illinois, and an affiliate hospital. Dietary intake was measured by using a 122-item version of the Block Health Habits and History Questionnaire, and correlations were assessed with Pearson's coefficients. Prostatic levels of tocopherols and carotenoids (but not retinol) were significantly correlated with plasma levels (r= 0.31-0.56, p < 0.05-0.0001); the strongest correlations were associated with lycopene, beta-carotene, and gamma-tocopherol (0.56, 0.54, and 0.52, respectively; p < 0.0001). Relative concentrations of tocopherols and carotenoids in prostate tissue were proportionate to those in plasma. No correlation between prostatic levels and reported dietary intake was observed (r = -0.09 to 0.16, p < not significant). Adjustment for energy intake, body mass index, and serum lipids did not impact these relations. These results suggest that plasma levels of tocopherols and carotenoids better reflect prostatic exposure than self-reported usual dietary intake.