The effective orifice area (EOA) is the standard parameter for the clinical assessment of aortic stenosis severity. It has been reported that EOA measured by Doppler echocardiography does not necessarily provide an accurate estimate of the cross-sectional area of the flow jet at the vena contracta, especially at low flow rates. The objective of this study was to test the validity of the Doppler-derived EOA. Triangular and circular orifice plates, funnels, and bioprosthetic valves were inserted into an in vitro aortic flow model and were studied under different physiologic flow rates corresponding to cardiac outputs varying from 1.5 to 7 L/min. For each experiment, the EOA was measured by Doppler and compared with the catheter-derived EOA and with the EOA derived from a theoretic formula. In bioprostheses, the geometric orifice area (GOA) was estimated from images acquired by high-speed video recording. There was no significant difference between the EOA derived from the 3 methods with the rigid orifices (Doppler vs catheter: y = 0.97x +0.18 mm(2), r(2) = 0.98; Doppler vs theory: y = 1.00x -3.60 mm(2), r(2) = 0.99). Doppler EOA was not significantly influenced by the flow rate in rigid orifices. As predicted by theory, the average contraction coefficient (EOA/GOA) was around 0.6 in the orifice plates and around 1.0 in the funnels. In the bioprosthetic valves, both EOA and GOA increased with increasing flow rate whereas contraction coefficient was almost constant with an average value of 0.99. There was also a very good concordance between EOA and GOA (y = 0.94x +0.05 mm(2), r(2) = 0.88). In rigid aortic stenosis, the Doppler EOA is much less flow dependent than generally assumed. Indeed, it depends mainly on the GOA and the inflow shape (flat vs funnel-shaped) of the stenosis. The flow dependence of Doppler EOA observed in clinical studies is likely a result of a variation of the valve GOA or of the valve inflow shape and not an inherent flow dependence of the EOA derived by the continuity equation.