Abstract According to meta-analyses aspirin provides a relative reduction in the rate of major vascular events of 19% in patients with arterial disease in general, whereas for patients with ischaemic cerebrovascular disease this reduction is only 13%. The discrepancy may well result from pathophysiological differences and not from a play of chance. There is no proven difference in efficacy according to dose. The evidence for this equivalence is most compelling in the range between 75 and 1300 mg daily, but still fairly convincing for doses between 30 and 50 mg. In contrast, side effects are clearly more frequent as the dose is higher. Other antiplatelet agents (sulfinpyrazone, ticlopidine, clopidogrel, dipyridamole, orally administered IIb/IIIa inhibitors) have no clear advantages over aspirin and in some cases definite disadvantages; the combination of aspirin and dipyridamole may be more efficacious than aspirin alone, but the evidence hinges on a single trial. If recurrent TIAs occur under treatment with aspirin, the rational response is not to change to a different antiplatelet agent, but to review the diagnosis and consider causes other than artery-to-artery embolism. Platelet aggregation can probably still occur despite complete acetylation of platelets, via pathways other than COX-1 inhibition, but in vitro aggregation tests are an unreliable measure.