Application of transcranial Doppler ultrasonography to asymptomatic prosthetic heart valve patients can result in the detection of transient high intensity signals, similar to those induced by the passage of emboli. However, the origin of these signals is unknown. An in vitro study has been undertaken to investigate the capacity of prosthetic heart valves to generate high intensity Doppler signals in the absence of blood. A pulse duplicator, filled with a seeded saline solution, was used to function prosthetic heart valves under mock-physiological conditions. A Björk-Shiley Monostrut valve was mounted in the aortic port while a tri-leaflet control valve was fixed in the mitral port. At stations upstream and downstream from the Björk-Shiley valve, flow was monitored using pulsed wave Doppler ultrasound (Nicolet TC-2000, 2 MHz probe). The effect of damping the harsh closure of the mechanical valve was investigated by applying a thin layer of soft adhesive tape between the valve occluder and outer ring. For all valve configurations, transient high intensity Doppler signals, characteristic of microemboli and similar to those observed in clinical studies of prosthetic heart valve patients, were detected downstream from the aortic port. The number of microemboli signals did not change significantly between sites at 20 cm and 40 cm downstream from the aortic valve. Damping the Björk-Shiley valve closure greatly reduced (by 80%) the number of microemboli signals detected. It is concluded that Doppler microemboli signals can be generated by prosthetic heart valves while functioning in the absence of the formed elements of blood, and that the number of microemboli signals produced depends upon the rate of energy dissipation at valve closure.