Initially, hydroxyethyl starch (HES) was only characterized by its in vitro molecular weight (MW). This is not sufficient because HES is degraded in vivo. One relevant parameter that predicts the rate of enzymatic breakdown is the degree of substitution, a measure of the average number of hydroxyethyl groups per glucose unit. The higher this degree of substitution, the slower the break-down. In addition, because the glucose units can be substituted at carbon 2, 3 and 6, different substitution patterns are possible. They are classified by their C2/C6 hydroxyethylation ratio. A higher C2/C6 ratio results in less metabolism of the starch in vivo and results in a larger in vivo MW. This in turn affects therapy, because the larger the in vivo MW, the longer is the duration of the volume effect of HES. Of particular importance is the fact that HES with a high in vivo MW affects factor VIII/von Willebrand factor which can lead to an acquired von Willebrand syndrome. During a 10-day volume therapy with a medium-MW HES 200, a form that is difficult to metabolize, we observed an 80% drop in factor VIII/von Willebrand factor. Therapy with a medium-MW HES 200, a form that is easily degraded, and therapy with a low-MW HES 70 did not result in a relevant decline of factor VIII/von Willebrand factor. This explains why hemorrhagic complications have been observed repeatedly in the United States after therapy with HES infusions, some of them lethal. In the United States high-MW HES 480 which is difficult to degrade is most frequently used and results in a larger in vivo MW and subsequent decrease in factor VIII/von Willebrand factor levels. In Europe, medium-MW HES 200 that is easily degraded and low-MW HES 70 are preferred. In the future, HES should be characterized by the in vivo, not the in vitro MW.