The pharmacokinetics of Factor VIII procoagulant activity (FVIII:C) were investigated in 10 patients receiving 2 different high purity concentrates of FVIII. Thein vivorecovery of FVIII:C depended on the method of blood volume estimation and on thein vitroassay used to define the potency of the concentrate; more generally, it also depends on the study design. The use ofin vivorecovery to compare FVIII preparations should therefore be discouraged and the achieved area under the FVIII activity-time curve (AUC) should be used instead.The volume of distribution at steady-state (Vss) of FVIII:C was significantly greater than the estimated plasma volume of the patients. The mean terminal half-life of FVIII:C was slightly assay-dependent, being 14h by a 1-stage clotting assay and 12h by a chromogenic assay. The calculated apparent clearance (CL) of FVIII:C, as well as the Vssand thein vivorecovery, depended on the measuredin vitropotency of the FVIII concentrate.These methodological aspects of pharmacokinetic studies are of direct interest to the clinician, since the routine administration of FVIII is based on the pharmacokinetic properties of the specific preparation.