Left heart bypass (LHBP) with heparin coated perfusion equipment including an arterial filter (pore size: 40 microns) was performed in five canine experiments after full systemic heparinization (heparin 300 IU/kg; activated coagulation time [ACT] > 480 sec). The heparin coated filter was replaced after 45 min with a second heparin coated filter. Protamine (1:1) was added after 45 min and perfusion was continued for another 45 min before the second filter was replaced with an uncoated control filter. In addition to continuous hemodynamic monitoring, all filters were disassembled and analyzed morphometrically with a scanning electron microscope (deposits on screens were expressed as percent of surface covered with fibrin or number of cells/100 micron 2, respectively). For the first heparin coated filter (ACT > 480), 0.3 +/- 0.5% of the surface was covered with fibrin, 0.7 +/- 0.7% with platelets, and 0.02 +/- 0.0% with red cells. For the second heparin coated filter exposed to neutralization of heparin with protamine, surface coverage was fibrin in 22 +/- 15%, platelets in 3.2 +/- 0.8%, and red cells in 0.2 +/- 0.1% (p < 0.05 for all comparisons with filter 1). For uncoated control filters (ACT = 120), surface coverage was fibrin in 31 +/- 33%, platelets in 3.7 +/- 2.9%, and red cells in 0.2 +/- 0.1% (not significant [NS] for all comparisons with filter 2). Although all arterial filters used in this study remained patent throughout the scheduled period, it became clear that protamine given during perfusion reduces the antithrombotic activity of bonded heparin. Hence, protaminization during perfusion with heparin coated equipment cannot be recommended.