Factor VII levels have been studied in hemophilia A and B plasmas and normal controls in a controlled, prospective study. Three assay methods were used: a standard clotting assay (FVIIc-A); a modified clotting assay (FVIIc-B) (Seligsohn et al, Blood 52:978-988, 1978); and a coupled amidolytic assay. By the FVIIc-B assay, the hemophilic plasmas were significantly lower than in the normal group (68.2 +/- 3.3% [SE] and 83.5 +/- 3.8%, respectively; P less than .01). The amidolytic assay, however, which measures total factor VII regardless of its activity state (factor VII or VIIa), was higher in the patient group than in the control group (126.9 +/- 9.6% and 99.4 +/- 5.7%, respectively; P less than .01). Control experiments showed that the differences in FVIIc-B activity were not caused by artifactual activation of factor VII ex vivo in the control group. The mean FVIIc-A assay of hemophilic plasmas (126.3 +/- 6.5%) agreed closely with the amidolytic assay, suggesting that the FVIIc-A method is also insensitive to the factor VII activity state. These data support the hypothesis that the FVIIc-B assay is more sensitive to the presence of factor VIIa. The increased sensitivity of the FVIIc-B assay to factor VII activation was confirmed by comparison of the two clotting assays on plasma subjected to activation in glass at 4 degrees C. The results of this study indicate that factor VII in hemophilic plasma is less activated than in normal plasma. Whether this contributes to the bleeding diathesis of hemophilia is unknown. However, it does provide evidence for the idea that factor VII in vivo is normally subject to some degree of activation by an enzyme (or enzymes) generated by a turnover of the intrinsic pathway.