Micronutrients, such as beta-carotene and vitamins A and E, are potential chemopreventive agents; however, their concentrations in human target tissues are largely unknown. Because these micronutrients may exert their action at the site of target tissues, the tissue concentrations of the micronutrients need to be determined. In this cross-sectional study, we have measured the concentrations of seven carotenoids, two retinoids, and two tocopherols in paired plasma, buccal mucosal cells (BMC), and skin samples from 96 healthy subjects (ages 26-82 yrs). The plasma-tissue, as well as the diet-plasma and diet-tissue relationships of the micronutrients, and the impact of various potential confounders on the micronutrient concentrations were evaluated. The micronutrient concentrations of plasma and BMC used in the evaluation were the average of three measurements over a one-month period. Our data indicated that 1) the correlations between the plasma and BMC (Spearman r = 0.40-0.91, p < 0.05) and the plasma and skin (r = 0.24-0.75, p < 0.05) concentrations of most micronutrients were significant in all subjects, suggesting that the status of these micronutrients in the BMC and skin may be estimated from their plasma concentrations; 2) the correlations between the diet and plasma/tissue concentrations of the micronutrients were generally not as strong as the plasma-tissue relationships; the diet-plasma and diet-tissue relationships of the carotenoids were particularly poor in the smokers; 3) the plasma and tissue concentrations of most micronutrients were lower in smokers than in nonsmokers and higher in vitamin supplement users than in nonsupplement users; the differences remained significant after adjustment for age, gender, and diet intake estimates; 4) among the seven carotenoids examined, lycopene was unique, because its concentration was not lower in smokers or higher in supplement users but was inversely associated with age.