Arterial stiffness is a measure of subclinical cardiovascular disease and increases with age. This study examines the association between arterial stiffness and cognitive decline in a cohort of older adults. A total of 2488 subjects with baseline measure of arterial stiffness (mean age, 74.2 years; 52.3% women) were prospectively followed over 9 years in the Health, Aging, and Body Composition Study. Arterial stiffness was measured as pulse wave velocity (PWV) and analyzed in tertiles. Cognitive function was assessed using the Modified Mini-Mental State examination at baseline and repeated at years 3, 5, 8, and 10. Lower Modified Mini-Mental State examination scores indicate worse function. We fit linear mixed models to examine longitudinal changes in cognitive function over the 9 years of follow-up and logistic regression models, restricted to 1331 participants, to examine cognitive impairment defined as a decrease of ≥5 points after 9 years. We adjusted for sociodemographics, Apoe4, and cardiovascular disease risk factors. The annual decrease in Modified Mini-Mental State examination scores was 0.30 points at low PWV (95% confidence interval [CI], -0.37 to -0.22), 0.46 points at middle PWV (95% CI, -0.54 to -0.39), and 0.45 points at high PWV (95% CI, -0.53 to -0.38), from fully adjusted linear mixed models. In fully adjusted models, the odds of cognitive impairment after 9 years of follow-up was 40% greater for subjects with middle PWV (odds ratio [OR], 1.40; 95% CI, 1.03-1.92) and 59% greater for subjects with high PWV (OR, 1.59; 95% CI, 1.16-2.18), compared with low PWV. High arterial stiffness was modestly associated with cognitive decline and impairment. Interventions to prevent arterial stiffness may be effective in delaying cognitive decline.