Factor VIII (FVIII) procoagulant activity was initially thought to be a glycoprotein with a molecular weight (MW) >1 million and composed of disulfide-1inked ~200,000 MW subunits. A protein with similar properties, except lacking procoagulant activity, is in hemophilic plasma; it was identical to normal FVIII by SDS-gel analyses, isoelectric focusing, and PAS staining. Subsequently it was shown that the FVIII glycoprotein also has von Willebrand factor (vWF) activity, suggesting that both FVIII and vWF activities might be properties of the same molecule. When the FVIII/vWF protein(s) is rechromatographed on 4% agarose and 0.25 M CaCl2, virtually all the protein and vWF activity elute in the void volume, but most of the FVIII procoagulant activity elutes much later. The extent of separation of the two activities depends on the amount of protein applied to the column. Also, exposure of the FVIII/vWF to thrombin before gel filtration strikingly accentuates separation of the two activities. The reduced SDS-gel pattern of the void volume protein peak showed the 200,000 MW subunit while that of the procoagulant peak contained several subunit bands which ranged from ~30,000–100,000 MW. Removal of sialic acid from FVIII/vWF is associated with reduced ristocetin induced platelet aggregation and causes a 50-fold increase in the rate of clearance of protein from the circulation by the hepatocyte. Currently, our data suggest that FVIII procoagulant and vWF activities are properties of a single molecule composed of disulfide-bound identical subunits. Cleavage by thrombin then results in FVIII procoagulant activity. Additional cleavages, to which the molecule appears very sensitive, results in FVIII inactivation. The vWF activity is very stable—even to proteolysis—and it appears to be a function of the carbohydrate side chains of the molecule.