An enzyme-linked differential antibody immunosorbent assay has been developed for the quantification of alpha2-plasmin inhibitor-plasmin and alpha2-macroglobulin-plasmin complexes. In this method the inhibitor-plasmin complex is bound to a surface by an inhibitor-specific antibody, and the plasmin bound to the inhibitor is quantified by a second antibody, rabbit antiplasminogen F(ab')2, labeled with alkaline phosphatase. The hydrolysis of p-nitrophenyl phosphate by the alkaline phosphatase is expressed in femtomoles of plasminogen per milliliter, by reference to a standard plasminogen curve. Inhibitor-enzyme complexes were generated in plasma by the addition of plasmin or of urokinase. The concentration of plasmin added was well below the plasma concentration of alpha2-plasmin inhibitor (1 microM) or of alpha2-macroglobulin (3.5 microM), so that neither inhibitor would be fully saturated with enzyme. Under these conditions increasing amounts of plasmin generated an increase in both alpha2-plasmin inhibitor-plasmin and alpha2-macroglobulin-plasmin complexes. Varying amounts of plasmin were incubated with each of the purified inhibitors in the concentration found in plasma, and the complexes. Varying amounts of plasmin were incubated with each of the purified inhibitors in the concentration found in plasma, and the complexes that formed were quantified by immunoassay. These studies made it possible to quantify the distribution of plasmin between the two inhibitors in plasmin or urokinase-treated plasma. In plasmin-treated plasma, 10% or less of the plasmin bound to both inhibitors was in complex with alpha2-macroglobulin. In contrast, between 19 and 51% of the plasmin generated in urokinase-activated plasma was bound to alpha2-macroglobulin. Thus, major changes in the distribution of plasma were observed, according to whether plasmin was added to plasma or whether plasminogen was activated endogenously. The pattern of inhibitor plasmin complexes generated in vivo by the therapeutic infusion of urokinase was similar to that found for urokinase-activated plasma. 23 normal individuals had low levels of alpha2-plasmin inhibitor-plasmin complexes, whereas six patients with laboratory evidence for disseminated intravascular coagulation demonstrated a 16- to 35-fold increase in he concentration of these complexes. These data indicated that a useful new probe for the study of the fibrinolytic enzyme system had been developed.