Randomized clinical trials provide the most reliable evidence of the risks and benefits of a treatment or management plan. Often it is necessary to have information from many thousands of patients in a clinical trial to know how to manage the next patient in the clinic. Recent trials in stroke prevention will change clinical practice by the more appropriate use of carotid endarterectomy and the use of anticoagulation in non-rheumatic atrial fibrillation. Other studies have indicated that modification of risk factors, such as decreased salt consumption, is important. Despite many new trials there is still no proven treatment for acute ischaemic stroke. To provide reliable evidence on the risks and benefits of treatment of acute stroke we need much larger randomized controlled trials.