The accuracy of right ventricular (RV) quantification by three-dimensional echocardiography (3DE) has been reported mainly in patients with a normal right ventricle (RV). However, there are no data regarding the accuracy of 3DE in patients with a dilated RV, as in shunt diseases. In this study, we evaluated the accuracy of 3DE and that of volumetric (Vol) cardiac magnetic resonance (CMR) for assessment of RV and left ventricular (LV) stroke volume (SV) and the pulmonary (Qp)/systemic (Qs) blood flow ratio in patients with an atrial septal defect (ASD) using the two-dimensional phase contrast (2DPC) method as the gold standard. We retrospectively investigated 83 patients with ASD who underwent transcatheter closure and clinically indicated CMR and 3DE examinations. The ratio Qp/Qs was calculated using RV and LV SV measured by full-volume volumetric 3DE (Vol-3DE) and CMR (Vol-CMR) and by two-dimensional pulsed Doppler quantification (2D-Dop); the parameters were compared using 2DPC-CMR as the gold standard. There was no significant difference in the Qp/Qs value between 2DPC-CMR and Vol-3DE (2.29±0.70 vs 2.21±0.63, P=.79) and 2D-Dop (vs 2.21±0.65, P=1.00); however, a significant difference was found between 2DPC-CMR and Vol-CMR (P<.001). The Qp/Qs value obtained using Vol-3DE showed the best correlation with 2DPC-CMR (r=0.93, P<.001). The RV and LV SV values obtained by Vol-3DE showed the best correlation with 2DPC-CMR (RV SV, r=0.82, P<.001; LV SV, r=0.73, P<.001), although the absolute values were underestimated. Qp/Qs was more accurately evaluated by Vol-3DE than by Vol-CMR or 2D-Dop. Three-dimensional echocardiography assessment was feasible and reproducible even in a dilated RV.