The Mayo Clinic medical records and records linkage system were used to identify stroke-free residents of Rochester, Minnesota, who were examined within 30 days after the first transient cerebral ischemic attack (TIA) during the period 1955 through 1979. The patients were divided into two groups: those given heparin within 30 days after the first attack and those not given heparin. Death, stroke, and either stroke or TIA were separate endpoints in Kaplan-Meier analyses of data from the day of initial examination through the 30th day thereafter. The probabilities of survival, survival free from stroke, and survival free from TIA for the heparin-treated group were not significantly different from those probabilities for the comparison group. The rate of hemorrhagic complications was 3.2 per 100 person-days of heparin therapy. Retroperitoneal hemorrhage, the most serious complication, was the cause of one death and one case of femoral neuropathy.