The published data of 34 studies of the long-term prognosis after transient ischaemic attacks (TIA) have been analysed. Despite the different types of investigation, the results, once the initial period of follow-up has been discounted, are consistent. The results indicate that, from approximately 1 year after onset of TIAs, the hazard rate is approximately constant and the level of excess annual mortality relative to general population experience is approximately 85%. At present, the data available to distinguish which is the superior between these two alternative descriptions of the long-term experience following TIA are not available.