The Windkessel model, coupled with the wave propagation theory, was applied to data measured in the ascending aorta of 11 anaesthetised dogs during total aortic occlusion at the thoracic and diaphragm levels. Wave speed and wave intensity were calculated using the measured pressure (P) and velocity (U), and separately using the pressure due to the wave (P(ex)) and U in the aorta approximately 1 cm distal to the aortic valve. Results show that wave speed, determined using the PU-loop method, is higher during thoracic than in diaphragm occlusion (p<0.001). On average wave speed calculated using P (c) is higher than that determined using P(ex) (c(WK)) in both occlusion sites (p<0.001). During aortic occlusion at the thoracic level, the intensity of backward waves was almost negligible using the Windkessel-wave system. Backward waves were observed during the occlusion at the diaphragm level, but their magnitude is lower compared to that determined with P. The Windkessel-wave system seems to reduce the magnitude of reflected waves during total aorta occlusion, notably if the occlusion sites are close to the ascending aorta.