More than 50 randomized trials have documented the efficacy and safety of aspirin as an antiplatelet agent and a cardiovascular drug. However, the optimal dose for preventing coronary and cerebral thrombosis has long been a cause of debate. For patients with ischaemic heart disease the range recommended for the prevention of a secondary event, based on strong clinical evidence, is 75-160 mg aspirin/day. For patients with cerebrovascular disease, recommendations range from 30-1300 mg/day. If these patients require a higher dose of aspirin it suggests that a different mechanism of action is involved. This paper considers hypotheses and reports the findings of recent clinical trials. The SALT study compared aspirin with placebo in 1360 patients with TIA or minor ischaemic stroke. It showed an 18% reduction in the risk of stroke or death in patients receiving 75 mg aspirin/day. Five other trials of 55,000 patients with ischaemic cerebrovascular disease compared the protective effect of aspirin (range 30-300 mg/day) with placebo, clopidogrel, or oral anticoagulants. Aspirin was better than placebo, safer than oral anticoagulants, and no different from clopidogrel. The implications of these findings are discussed. Mechanistic studies and randomized clinical trials strongly suggest that the mechanism of action and dose requirement of the antithrombotic effect of aspirin in patients with cerebrovascular disease is the same as that for ischaemic heart disease.