Murine models of left ventricular (LV) hypertrophy recently have been developed. We tested the accuracy of 2-dimensional (2D) echocardiographic measurement of LV mass with high-frequency imaging in mice. Ten anesthetized mice (weight 20 to 31 g, aged 1 to 5 months) were examined with a 15-MHz transthoracic linear-array transducer. End-diastolic myocardial area (A)(epicardial - endocardial) from the parasternal short-axis view at the midpapillary level and LV length (L) from the parasternal long-axis view were measured to calculate LV mass with the area-length method (1.05 [5/6 x A x L]) and data were compared with LV-mass with the 2D guided M-mode method. Within 3 days of echocardiography, the hearts were removed and weighed after potassium-induced cardiac arrest. Two-dimensional echocardiographic measurement with a 15-MHz transducer was performed in all mice. LV chamber dimensions included end-diastolic septal (0.80 ± 0.12 mm) and posterior wall thickness (0.76 ± 0.13 ram), end-diastolic dimension (3.64 ± 0.28 mm), and end-systolic dimension (2.34 ± 0.32 mm). Echocardiographic LV mass with the area-length method, 2D guided M-mode method, and autopsy LV weight were 80.8 ± 16.1 mg, 97.6 ± 17.8 mg, and 78.8 ± 13.2 mg, respectively. A strong correlation existed between LV weight (x) and echocardiographic LV mass (y) with the area-length method: y = 0.745x + 18.9, r =0.908, standard error of estimate (SEE) = 5.9 mg, P < .0005. This correlation was stronger than that of LV weight (x) and echocardiographic LV mass (y) with the 2D guided M-mode method: y = 0.577x + 22.6, r =0.779, SEE = 8.8 mg, P = .008. These data suggest that serial in vivo measurements of LV mass with the 2D area-length method may be more accurate than M-mode methods in experimental murine models of LV pathology.