A possible association between body iron stores, measured as serum ferritin, and carotid arterial intima-media thickening was investigated in the Atherosclerosis Risk in Communities Study during 1990-1992 using a matched case-control design. For a 143-micrograms/liter greater serum ferritin concentration (the interquartile range), the odds ratio for cases with carotid intima-media thickening versus controls was 1.12 (95% confidence interval 0.97-1.30). However, there was no association (odds ratio = 1.00) after adjusting for major cardiovascular risk factors. This analysis of carotid arterial intima-media thickening, a measure of early atherosclerosis, in relation to serum ferritin does not support the hypothesis that increased body iron stores increase the risk of atherosclerotic cardiovascular disease.