Intentional hemodilution has been shown to increase cerebral blood flow under normal conditions and in experimental stroke. Clinical trials of hemodilution in patients with acute stroke have employed dextran, hydroxyethyl starch or albumin as hemodiluting agents, often in combination with venesection. An overview of 12 randomized clinical trials of hemodilution in acute stroke (total 2,509 patients) shows no overall effect on survival (odds ratio 1.01 for active treatment:control). In surviving patients, neurological outcome is similar in hemodilution and control groups. The proportion of patients that are independent at follow-up does not differ between the groups (odds ratio 0.98). It has not been possible to identify any subgroup in which the therapy is clearly beneficial. Recent pilot clinical studies indicate that the effects of more aggressive hemodilution schemes could be worthwhile to explore. Until the results of new controlled trials are available, it seems that hemodilution has no place in the routine management of patients with ischemic stroke.