During the development of cardiovascular devices, it is necessary to evaluate their performance in vitro under experimental conditions that closely resemble their use in vivo. Therefore, an in vitro test model was developed to evaluate the performance of cardiovascular devices with blood. The in vitro test model consists of a pneumatically driven 50 cc polyurethane ventricle with either mechanical or experimental polyurethane tricuspid semilunar valves, connected to a horseshoe-shaped blood reservoir. A sampling port permits frequent removal of blood for hematologic evaluation. All the couplings are of the quick connect type to facilitate assembly and exchange of parts. The total volume of blood in the test model is approximately 1.4 L, and experiments are conducted in the absence of an air-blood interface. Except for mechanical valves and the chosen test interface, the entire blood contacting surface consists of polyurethane. Fresh bovine heparinized blood is used for the experiments. After filling the reservoir and the ventricle with blood, the concentration of heparin is adjusted with protamine to an APTT level of three times normal (80-100 sec). Once this level of anticoagulation is achieved, the blood is circulated for the next 2-3 hr at the desired test parameters (beats/min, cardiac output, etc.). A number of hematologic parameters are monitored during the test, including APTT, PT, plasma free hemoglobin and ADP-induced platelet aggregation. With the help of this model a number of devices, including experimental polyurethane tricuspid semilunar valves, Silastic tube valves, valves with heparin coated leaflets, and small (50 cc) experimental ventricles have been studied.(ABSTRACT TRUNCATED AT 250 WORDS)