Résumé : BACKGROUND AND PURPOSE - The widespread preference of anticoagulants over antiplatelets in patients with cervical artery dissection (CAD) is empirical rather than evidence-based. SUMMARY OF REVIEW - This article summarizes pathophysiological considerations, clinical experiences, and the findings of a systematic metaanalysis about antithrombotic agents in CAD patients. As a result, there are several putative arguments in favor as well as against immediate anticoagulation in CAD patients. CONCLUSIONS - A randomized controlled trial comparing antiplatelets with anticoagulation is needed and ethically justified. However, attributable to the large sample size which is required to gather meaningful results, such a trial represents a huge venture. This comprehensive overview may be helpful for the design and the promotion of such a trial. In addition, it could be used to encourage both participation of centers and randomization of CAD patients. Alternatively, antithrombotic treatment decisions can be customized based on clinical and paraclinical characteristics of individual CAD patients. Stroke severity with National Institutes of Health Stroke Scale score ≥15, accompanying intracranial dissection, local compression syndromes without ischemic events, or concomitant diseases with increased bleeding risk are features in which antiplatelets seem preferable. In turn, in CAD patients with (pseudo)occlusion of the dissected artery, high intensity transient signals in transcranial ultrasound studies despite (dual) antiplatelets, multiple ischemic events in the same circulation, or with free-floating thrombus immediate anticoagulation is favored. © 2007 American Heart Association, Inc.