The major treatable risk factors in thromboembolic stroke are hypertension and transient ischemic attacks (TIA). Two prospective hospital series of TIA provide no evidence that either medical or surgical therapy is superior to the other or to supportive treatment. However, the treated groups were not shown to be comparable. Only the joint cooperative study on extracranial arterial occlusion provided random allocation of TIA patients to surgery or no surgery. The published conclusion comparing course after discharge was that surgery was of significant benefit. However, when the groups were compared from the point of randomization and when later TIA alone were excluded, there was not even a suggestion that the surgery groups did better than the others. Sample size was such, though, that a possible reduction of stroke occurrence by one-half could not be excluded with the data at hand, leaving the ultimate decision as to the potential value of surgery in TIA to the clinician.