In a random manner, 100 patients with acute cerebral infarction due to thromboembolism were assigned to two groups. The patients in one group received continuous intravenous therapy with dextran 40 for three days; the others received an equivalent volume of fluid without the dextran 40. In the treated group, levels of dextran 40 remained greater than 1% for three to five days after the infusion was stopped. The group treated with dextran 40 showed lower mortality and better quality of survival in terms of the neurological status. The improvement in the treated group was significantly better than in the untreated group (P<0.05). Platelet appearance in those with acute stroke showed abnormalities compared to normals, and platelet aggregation in venous blood was increased. Dextran 40 in vitro and in vivo tended to return platelet features to normal.