Two hundred eighty-four female adults (aged 40-70 years) were longitudinally studied to investigate the relationship between dietary supplemental vitamin A and serum biochemical markers of vitamin A toxicity. Serum retinol, retinyl esters, and retinol-binding protein (RBP), alkaline phosphatase and aspartate aminotransferase activities and bile acids were measured at baseline, 1 and 2 years. Fasting serum retinol and retinyl ester concentrations were determined by high-performance liquid chromatography, and dietary and supplemental intake of vitamin A were assessed by 3-day food records. There was no difference in dietary vitamin A intake between supplement users and nonusers. In supplemental users, the mean +/- SEM supplemental vitamin A intake was 952 +/- 81 IU/day (range 250-5000 retinol equivalents/day). Serum retinol, retinyl esters, and RBP concentrations were not different between the two groups during the 2-year period. For each group, serum retinyl esters significantly increased over time (p < 0.03), but the magnitude of the increase was not different between the groups. Serum levels of retinol, retinyl esters, and RBP were not correlated with vitamin A intake or age in either group. Biochemical measures of liver damage (serum alkaline phosphatase and aspartate aminotransferase activities and serum bile acids) were not related to serum retinol, retinyl esters or RBP concentrations, nor were they different between nonusers and users of supplemental vitamin A. This study provides evidence that long-term supplemental vitamin A in doses commonly found in multivitamin supplements does not present a risk for hypervitaminosis A.