After extensive experimental evaluation, heparin-coated perfusion equipment was clinically evaluated with low or no systemic heparinization in three different groups of patients (n = 47). In group 1, resection of descending thoracic aortic aneurysms (n = 24) was performed with heparin-coated equipment used for left heart bypass (n = 12) or partial cardiopulmonary bypass (n = 12) for proximal unloading and distal protection (heparin 5000 IU, autotransfusion). All devices remained functional throughout the procedures and no systemic emboli were detected. The sole death (1 of 24, 4%) occurred in a patient with ruptured thoracoabdominal aortic aneurysm requiring operation in extremis. Paraparesis with spontaneous recovery occurred in one patient (1 of 24, 4%). In group 2, coronary artery revascularization randomized for low (activated clotting time greater than 180 seconds) versus full (activated clotting time greater than 480 seconds) systemic heparinization was prospectively analyzed in 22 patients. All patients recovered without sequelae, and no myocardial infarction was diagnosed. Low dose of heparin (8041 +/- 1270 IU versus 52,500 +/- 17,100 IU; p less than 0.0005) resulted in reduced protamine requirements (7875 +/- 1918 IU versus 31,400 +/- 14,000 IU; p less than 0.0005), reduced blood loss (831 +/- 373 ml versus 2345 +/- 1815 ml; p less than 0.01), reduced transfusion requirements of homologous blood products (281 +/- 415 ml versus 2731 +/- 2258 ml; p less than 0.001), and less patients transfused (5 of 12 versus 10 of 10; p less than 0.05). Lower D-dimer levels in the group perfused with low systemic heparinization (0.50 +/- 0.43 mg/L versus 1.08 +/- 0.59 mg/L; p less than 0.05) were attributed to the absence of cardiotomy suction in this group. In group 3, rewarming in accidental hypothermia by cardiopulmonary bypass was successfully performed without systemic heparinization in a patient with hypothermic cardiac arrest (23.3 degrees C) and intracranial trauma. We conclude that systemic heparinization for clinical cardiopulmonary bypass can be reduced and eliminated in selected patients if perfusion equipment with improved biocompatibility is used. Bypass-induced morbidity can be reduced.