Prevention of valve thrombosis in patients after prosthetic mechanical heart valve replacement and heparin-induced thrombocytopenia (HIT) is still an open issue. The aim of the present in-vitro study was to investigate the efficacy of argatroban and bivalirudin in comparison to unfractionated heparin (UFH) in preventing thrombus formation on mechanical heart valves. Blood (230 ml) from healthy young male volunteers was anticoagulated either by UFH, argatroban bolus, argatroban bolus plus continuous infusion, bivalirudin bolus, or bivalirudin bolus plus continuous infusion. Valve prostheses were placed in a newly developed in-vitro thrombosis tester and exposed to the anticoagulated blood samples. To quantify the thrombi, electron microscopy was performed, and each valve was weighed before and after the experiment. Mean thrombus weight in group 1 (UFH) was 117 + 93 mg, in group 2 (argatroban bolus) 722 + 428 mg, in group 3 (bivalirudin bolus) 758 + 323 mg, in group 4 (argatroban bolus plus continuous infusion) 162 + 98 mg, and in group 5 (bivalirudin bolus plus continuous infusion) 166 + 141 mg (p-value <0.001). Electron microscopy showed increased rates of thrombus formation in groups 2 and 3. Argatroban and bivalirudin were as effective as UFH in preventing thrombus formation on valve prostheses in our in-vitro investigation when they were administered continuously. We hypothesise that continuous infusion of argatroban or bivalirudin are optimal treatment options for patients with HIT after mechanical heart valve replacement for adapting oral to parenteral anticoagulation or vice versa.