Recently, we and other groups showed that analysis of transthoracic pulmonary venous flow velocity patterns is useful in the estimation of left ventricular diastolic function and left atrial pressure; however, it is unclear whether transthoracic tracings, in contrast to transesophageal tracings, provide reliable quantitation of the abnormal pulmonary venous flow characteristics in patients with and without cardiac disorders. This study was attempted to validate transthoracic pulmonary venous flow measurements by comparing them with transesophageal measurements in eight normal volunteers and 17 patients with a variety of cardiac disorders. Transesophageal and transthoracic measurements of peak diastolic forward flow velocity, the ratio of peak systolic/diastolic forward flow velocities, and the peak reversal flow velocity at atrial contraction sampled in the right upper pulmonary vein showed good correspondence to each other. The flow velocities were higher in the right upper pulmonary vein than in the left upper pulmonary vein. Narrower spectral flow velocity patterns with higher velocities were obtained 1 to 2 cm distal to the orifice in the pulmonary vein than at the orifice. Thus transthoracic measurements of the pulmonary venous flow velocity pattern are feasible and accurate in patients and may be used to assess left ventricular function and hemodynamics as a substitute for transesophageal measurements.