In a cross-sectional survey of 86 elderly persons, it was observed that subjects with elevated plasma alpha-tocopherol levels had depressed plasma gamma-tocopherol. Tocopherols were measured by both reverse-phase and normal-phase high performance liquid chromatography (HPLC). When eight human volunteers (age range 30-60) were given 1200 IU of all-rac-alpha-tocopherol daily for 8 wk, plasma gamma-tocopherol and beta-tocopherol decreased in all subjects. After supplementation, gamma-tocopherol values were typically 30-50% of initial values, and alpha-tocopherol values were typically 200-400% of initial values. These results suggest that intestinal uptake and/or plasma transport make more efficient use of alpha-tocopherol than of gamma- or beta-tocopherol. Moreover, the results indicate that the ratio of gamma- to alpha-tocopherol in plasma would be a more satisfactory index to measure compliance in trials involving supplementation with alpha-tocopherol.