Cardiovascular risk prediction relies on classical risk factors such as age, gender, lipids, hypertension, smoking and diabetes. Although the value of such scales of risk is high for populations, its value for individual is reduced and too much influenced by non-modifiable risk factors (age and gender). Biomarkers of risk have been deceiving and genome wide scan approach is too recent. Target organ damage may help in selecting patients at high risk and in determining intervention. Aortic pulse wave velocity, an index of aortic stiffness, has been widely validated as providing additional risk predictions beyond and above classical risk factors, and has now entered into official guidelines. Many interventions (dietary, behaviour, drug treatment) were shown to influence arterial stiffness positively, but little evidence of a direct effect of intervention on arterial stiffness independent of blood pressure is available. New pharmacological targets and new drugs need to be identified. To become a surrogate endpoint for drug development, there is a need to demonstrate that regression arterial stiffness is associated with improved outcome. In parallel to this demonstration, points to be improved are the homogenization and spreading of the technique of measurement, the establishment of a reference value database.