The aortic-femoral arterial stiffness gradient, calculated as the ratio of lower-limb pulse-wave velocity (PWV) to central (aortic) PWV, is a promising tool for assessing cardiovascular disease (CVD) risk, but whether it predicts incident CVD is unknown. We examined the association of the aortic-femoral arterial stiffness gradient measures carotid-femoral stiffness gradient (femoral-ankle PWV divided by carotid-femoral PWV) and the heart-femoral stiffness gradient (femoral-ankle PWV divided by heart-femoral PWV), as well as PWV, with incident CVD (coronary disease, stroke, and heart failure) and all-cause mortality among 3109 participants of the Atherosclerosis Risk in Communities Study cohort (age, 75±5 years; carotid-femoral PWV, 11.5±3.0 m/s), free of CVD. Cox regression was used to estimate hazard ratios (HR) and 95% CIs. Over a median 7.4-year follow-up, there were 322 cases of incident CVD and 410 deaths. In fully adjusted models, only top quartiles of carotid-femoral stiffness gradient (quartile 4: HR, 1.43 [95% CI, 1.03-1.97]; and quartile 3: HR, 1.49 [95% CI, 1.08-2.05]) and heart-femoral stiffness gradient (quartile 4: HR, 1.77 [95% CI, 1.27-2.48]; and quartile 3: HR, 1.41 [95% CI, 1.00-2.00]) were significantly associated with a greater risk of incident CVD. Only high aortic stiffness in combination with low lower-limb stiffness was significantly associated with incident CVD (HR, 1.46 [95% CI, 1.06-2.02]) compared with the referent low aortic stiffness and high lower-limb stiffness. No PWVs were significantly associated with incident CVD. No exposures were associated with all-cause mortality. The aortic-femoral arterial stiffness gradient may enhance CVD risk assessment in older adults in whom the predictive capacity of traditional risk factors and PWV are attenuated.