We compared survival following transient ischemic attack (TIA) in 2 prospective cohorts of TIA patients admitted to Wake Forest University Medical Center. The 1st consisted of 177 patients admitted between 1961 and 1973, and the 2nd of 185 patients admitted between 1980 and 1983. Patients in the 2nd cohort had significantly greater longevity than patienta in the 1st cohort, both univariately and after adjustment for cerebrovascular risk factors. The adjusted 1-year survival estimate increased from 91% in the 1st cohort to 98% in the 2nd, and the adjusted 3-year survival estimate increased from 83% in the 1st to 94% in the 2nd. The underlying causes for this dramatic improvement in survival may include early identification and aggressive management of TIAs or coexisting diseases, improved management of subsequent completed strokes or myocardial infarctions, or unadjusted differences in these cohorts. The data imply that reports of TIA survival from different periods may not be comparable. © 1989 American Academy of Neurology.