I n a previous paper, we showed that the abnormality of ristocetin-induced platelet aggregation in platelet-rich plasma in 10 patients with von Willebrand's disease could be corrected by a factor in normal plasma that was present in the same fractions as factor VIII procoagulant activity (antihemophilic factor, AHF, VIII_AHF) when prepared by chromatography on Bio-Gel 5 M (Bio-Rad Laboratories, Richmond, Calif.). This observation suggests that patients with this disorder are deficient in a plasma factor, associated with the factor VIII molecule, that is necessary for normal platelet function. In the present paper, we describe, an assay for this factor, the von Willebrand factor (VIII_VWF), based on the observation that a log-log relationship exists between the amount of ristocetin-induced aggregation of washed, normal platelets and the concentration of normal plasma present in the test system. We assayed the activity of VIII_VWF as well as antihemophilic factor procoagulant activity (VIII_AHF) and factor VIII antigen (VIII_AGN) in 15 patients with von Willebrand's disease and 20 normal subjects. A highly significant correlation (r ∼ 0.80) between VIII_VWF and both VIII_AHF was found in normal subjects and in patients with von Willebrand's disease. This finding, in addition to the observation that agarose gel chromatography fractions that have VIII_AHF procoagulant activity also have VIII_VWF activity, strongly suggests that the von Willebrand factor is associated with the factor VIII molecule. VIII_VWF in normal plasma was not inhibited by human anti-VIII, and VIII_VWF levels were normal in hemophilic plasma. Thus, the VIII_VWF site on the factor VIII molecule appears to be different from that determining VIII_AHF. Finally, the activity of VIII_VWF appeared to correlate better with the bleeding time than either VIII_AHF or VIII_AGN. This suggests that VIII_VWF assayed in this study may be the “anti-bleeding factor” that is deficient in von Willebrand's disease. These findings are consistent with a decreased synthesis of the factor VIII molecule in von Willebrand's disease and suggest the possibility of additional abnormalities of the site on the molecule that determines the activity of VIII_VWF.