Incubation of a factor VIII-rich fraction of plasma with a high concentration of salt confirmed the production of both high (HMW) and low (LMW) molecular weight factor VIII clotting activity (FVIIIC) as determined by agarose gel filtration but with considerable overlap. The electrophoretic mobility of factor VIII related protein (FVIIIRP) detected by precipitating rabbit antiserum was not affected by this treatment and LMW FVIIIC devoid of FVIIIRP was apparently produced. At low concentration the reducing agent dithiothreitol (DTT) altered the electrophoretic mobility of FVIIIRP. At higher concentrations it altered both its mobility and antigenicity and an LMW FVIIIRP was produced. Contrary to the findings of other workers no LMW FVIIIC devoid of FVIIIRP was produced. In further studies factor VIII-rich plasma fraction was treated with sepharose beads to which had been coupled a non-coagulation inhibitory precipitating rabbit antibody to FVIIIRP. Both FVIIIRP and FVIIIC were taken up by the beads but after elution with 1.5 M NaCl, FVIIIC of LMW and devoid of FVIIIRP was selectively removed. Antisera raised to LMW FVIIIC produced with 1.5 M NaCl either by the gel filtration or affinity chromatography methods inhibited FVIIIC and precipitated with HMW factor VIII-rich fractions. The results were consistent with the possibility that factor VIII clotting activity and FVIIIRP exist in plasma as a non-covalently bound complex.