Previous cross-sectional studies have demonstrated significant associations between the ankle-brachial index (ABI) and carotid intimal medial thickness (IMT). In a longitudinal study, 637 participants who had ABI measurements from 1992 to 1994 were tested for carotid IMT an average of 4.8 years later. Mean age was 72.5 years; 60% were women. More than half (51%) were hypertensive, whereas 28% had dyslipidemia, 56% were past or current smokers, and 6% had diabetes. The mean +/- SD ABI and IMT were 1.06 +/- 0.13 and 1.28 +/- 0.44 mm, respectively. Mean age- and gender-adjusted IMTs for the ABI groups of <or=0.90, 0.91 to 0.99, and 1.00 to 1.09 were significantly greater than those from 1.10 to 1.26 (1.53, 1.49, 1.28, and 1.18 mm, respectively, p <0.05 for all). In logistic regression analyses, each of these ABI groups was significantly associated with increased odds for an IMT in the highest decile compared with an ABI of 1.10 to 1.26 (odds ratios 3.2, 4.3, and 2.3, respectively, p <0.05 for all). In conclusion, an ABI of 1.10 to 1.26 was associated with the lowest levels of cardiovascular disease risk factors and carotid atherosclerosis, suggesting that a new cut point for ABI risk stratification should be considered.