Assessment of right ventricular volume and function is important in many clinical settings involving heart or lung disease. However, the complexity of the right ventricular anatomy has prevented accurate volume determination by two-dimensional echocardiography. In the present study, 5 models incorporating standard echocardiographic views, were used to determine right ventricular volume in 10 human subjects. Two models were contingent on the true crescentic appearance of the right ventricle, whereas the remaining 3 calculated the right ventricular volume as a pyramid, an ellipsoid or other tapering geometrical figures, respectively. Subsequently, echocardiographic right ventricular volumes were compared to magnetic resonance imaging derived volumes. Correlation analysis and agreement measurement between the echocardiographic and magnetic resonance end-diastolic volume were performed in 10 out of 10 subjects and in 9 out of 10 subjects for the end-systolic volume. The 2 crescentic models resulted in the most reliable estimation of right ventricular volume. Those findings suggest that models based on right ventricular anatomical landmarks are feasible and should be preferred in echocardiographic studies.