Although carotid intima-media thickness (IMT) has been broadly used as a tool to evaluate cardiovascular risk, its role as a surrogate endpoint is still debated. The main issue is the fact that no study has ever been powered to show a relationship between changes in carotid IMT during follow-up and cardiovascular events. A meta-analysis of existing clinical studies was performed to investigate this relationship but it failed to demonstrate a predictive role of regression in carotid IMT for cardiovascular events. The reasons for the lack of a clear evidence for a predictive role of IMT progression are unknown but are likely multifactorial. Firstly, it may depend on the fact that this index is not a pure atherosclerosis index. Second, carotid atherosclerosis does not always reflect coronary atherosclerosis. Furthermore, methodologic problems related to intra- and interobserver variability make this index not adequately reproducible when tracking the progression of carotid atherosclerosis. A further meta-analysis based on individual patient data, instead of published data, has been planned to better address the predictive role of IMT. Lastly, in the future, the variability of ultrasound measurements of carotid IMT are likely to be reduced by further development of automatic calculation of this index by magnetic resonance imaging.