Background-Real-time 3-dimensional echocardiography (RT3DE) is superior to 2D echocardiography in assessing left ventricular (LV) volumes and ejection fraction (EF), but its feasibility is limited by multibeat acquisition, which requires an optimal breath-hold and a regular heart rhythm. We sought to evaluate the accuracy and feasibility of single-and 2-beat RT3DE for LV volume and EF assessment. Methods and Results-Sixty-six consecutive patients referred for cardiac magnetic resonance (CMR) underwent RT3DEand CMR on the same day. Of the 50 patients (age, 59Ϯ18 years; 68% men; 42% coronary artery disease; LVEFϭ49Ϯ14%; limits, 14% to 76%) with an adequate RT3DE image quality, accuracy for LV volumes and EF measurements of single-and 2-beat modalities were compared with the conventional 4-beat acquisition and CMR.Correlations with CMR for LV end-diastolic volume (161Ϯ59 mL, rϭ0.93 to 0.94) and end-systolic volume (86Ϯ56 mL, rϭ0.93 to 0.96) were excellent regardless of the number of cardiac cycles used. However, because of the low temporal resolution (7Ϯ2 volumes per second), single-beat underestimated LVEF (bias, Ϫ5Ϯ8%) with greater bias than 2-beat (bias, 1Ϯ6%, PϽ0.001) and 4-beat (bias, 3Ϯ7%, PϽ0.001) modalities. Interestingly, 2-beat provided accuracy similar to 4-beat for end-diastolic volume (bias, Ϫ17Ϯ21 mL versus Ϫ15Ϯ23 mL), end-systolic volume (bias, Ϫ9Ϯ16 mL versus Ϫ12Ϯ17 mL), and LVEF (bias, 1Ϯ6% versus 3Ϯ7%) measurements, but fewer stitching artifacts were observed with 2-than 4-beat modalities (3% versus 30%). Conclusions-Compared with conventional multibeat acquisitions, 2-beat modality provides similar accuracy in LV volume and EF measurements and should be preferred due to fewer stitching artifacts. In contrast, the temporal resolution of single-beat modality appears insufficient to provide an accurate estimation of LVEF. (Circ Cardiovasc Imaging. 2010;3:450-455.)