Binding of plasmin (ogen) to rat C6 glioma cells is saturable and kringle-domain dependent. This interaction was studied as a model of plasimin(ogen) receptor interactions in nucleated mammalian cells. Apparentl25I-plasmin dissociation from OS cell binding sites was slow; however, the dissociation rate was increased when the solution contained diisopropyl phosphoryl-plasmin (0.3 juM), fibrinogen (0.16 or 0.8 mg/ml), 1.08 mM D-Val-L-Leu-L-Lys-p-nitroanilide-HCl (S-2251), or e-amino-n-caproic acid (EACA, 5.0 mM). EACA promoted the most rapid dissociation of plasmin. C6 cell-associated plasmin and plasmin in solution demonstrated similar amidase activity. Only specifically bound plasmin (75% of total binding) was active against S-2251. Plasmin that was initially bound to C6 cells digested fibrinogen in a timeand plasmin concentration-dependent manner. α-Antiplasmin (α2AP, 0.1fiM)completely inhibited fibrinogenolysis by plasmin that was initially C6 • or human umbilical vein endothelial-cell associated. Sincea2APreacts selectively with plasmin in solution (minimally with plasmin bound to cells), fibrinogen digestion by cell-associated plasmin probably occurred only after the plasmin dissociated into solution. Crosslinked fibrin clots were formed in uniform layers over C6 cells. If the cells were incubated with plasmin before addition of fibrinogen and thrombin, the clots were rapidly lysed. α2AP incompletely inhibited fibrinolysis when added after fibrin polymerization (44% inhibition with 0.1 μM or α2AP). Fibrinolysis was completely inhibited when t$AP was added before fibrin polymerization. These studies suggest that plasmin must first dissociate from cellular binding sites to mediate fibrinogenolysis or fibrinolysis. After dissociation, plasmin activity is modulated by antiplasmins, which are ineffective at the cell surface.