Left atrial (LA) size is an important predictor of cardiovascular events. Various methods of LA volume assessment exist, but their differences have not been defined. This prospective study included 631 patients (331 men; mean age of 68 ± 14 years) without a history of atrial arrhythmias, stroke, valvular heart disease, pacemaker implantation, or congenital heart disease. All underwent echocardiography with comprehensive diastolic function assessment and LA volume measurement by 3 commonly used methods: biplane area–length, biplane Simpson’s method, and the prolate-ellipsoid method. Mean LA volumes were 39 ± 14 ml/m² by the area–length method, 38 ± 13 ml/m² by the Simpson’s method, and 32 ± 14 ml/m² by the prolate-ellipsoid method. In 92% of patients, the prolate measurement was smaller than the 2 biplane methods. Pairwise correlations (r) were 0.98 for area–length versus Simpson’s, 0.85 for prolate versus area–length, and 0.86 for prolate versus Simpson’s (all p values <0.001). For distinguishing normal (n = 62) from pseudonormal diastolic function (n = 240) using receiver-operating curve analysis, areas under the curves were 0.76, 0.78, and 0.75 for the area–length, Simpson’s, and prolate methods, respectively (all p values <0.001, no significant intermethod differences). In conclusion, our findings suggest that there are systematic differences among existing LA volume methods. Biplane area–length and Simpson’s methods compare closely, whereas the prolate-ellipsoid method generally yields smaller volumes.