The use of experimental models of arterial thrombosis both in vivo and ex vivo in animals and ex vivo in humans is an obligatory step for the understanding of mechanisms involved in thrombogenesis as well as in the evaluation of anti-thrombotic therapeutics. Arterial thrombogenesis is a complex phenomenon which involves multiple systems, mechanisms and parameters. Therefore studies of thrombogenesis from a pathological as well as a therapeutic point are necessary for understanding this problem in its entirety. For these reasons, it is necessary to use models as representative as possible of the human pathological condition. Besides these theoretical requirements, practical needs have also to be fulfilled (accessibility of the models, adaptation to the type of the technique to different animal model and/or of the size of the animal to the amount of molecule available, cost...) which necessarily lead to some compromises. In this review we have tried to underline the criteria for the choice, characteristics, advantages and disadvantages of the major models commonly accepted and used, in such a form that the reader who may not be an expert in the field would be led either to a choice of a particular model for a specific purpose or to appreciate a paper or a report based on an experimental model of arterial thrombosis. In vitro models of arterial thrombosis are so far removed from reality and due to their nature can generate so much artifacts that we have omitted their discussion from this paper.