Several investigations have shown that factor VIII is a high molecular weight aggregate and that both non-covalent and disulphide bonds contribute to the stabilization of the macromolecular factor VIII aggregate. As shown previously (1) disruption of non-covalent bonds can be accomplished at relatively low ionic strength and neutral pH and results via a series of homologous oligomers (having constant charge/mass ratios) in 2 immunologically non-related subunits. Thus, the generally accepted concept that factor VIII is constructed of identical subunits seems incorrect. For several reasons we assume that the observed fragmentation at low ionic strength is not due to proteolytic breakdown. However, this view is not favoured by the observation that tryptic and plasmic digestion of factor VIII (in aggregated form) results in gel electrophoresis patterns comparable with those obtained of factor VIII after low ionic strength dissociation. Further, evidence will be presented that the assumption that reduction of factor VIII under denaturating conditions results in a single polypeptide chain is also no longer tenable. As far as the reduction of aggregated factor VIII is concerned our observations agree with data from the literature, that is, reduction of factor VIII results in 1 major subunit with an apparent molecular weight of approximately 270.000. However, when factor VIII is first dissociated at low ionic strength followed by reduction, smaller fragments appear in the electrophoresis pattern. Thus, it seems most likely that the apparent single chain subunit is in fact, constructed of smaller non-covalently linked fragments.(1) J. A. van Mourik et al. Thrombosis Research, 4, 155, 1974.