BACKGROUND: Evaluation of flow convergence (FC) with two-dimensional (2D) imaging systems may not be sufficiently accurate to characterize these often asymmetric, complex phenomena. The aim of this study was to validate a three-dimensional (3D) method for determining the severity of aortic regurgitation (AR) in an experimental animal model. METHODS AND RESULTS: In six sheep with surgically induced chronic AR, 20 hemodynamically different states were studied. Instantaneous regurgitant flow rates were obtained by aortic and pulmonary electromagnetic flow meters. Video composite data of color Doppler flow mapping images were transferred into a TomTec computer after computer-controlled 180 degrees rotational acquisition. Direct measurement of the 3D reconstructed FC surface areas as well as measurements of FC areas estimated with 2D methods with hemispherical and hemielliptical assumptions were performed, and values were multiplied by the aliasing velocity to obtain peak regurgitant flow rates. There was better agreement between 3D and electromagnetically derived flow rates than there was between the 2D and the reference values (r=.94, y=1.0x-0.16, difference=0.02 L/min for the 3D method; r=.80, y=1.6x-0.3, difference=1.2 L/min for the 2D hemispherical method; r=.75, y=0.90x+0.2, difference=-0.20 L/min for the 2D hemielliptical method). CONCLUSIONS: Without any geometrical assumption, the 3D method provided better delineation of the FC zones and direct measurements of FC surface areas, permitting more accurate quantification of the severity of AR than the 2D methods.