OBJECTIVE This study was designed to assess whether the transmitral flow velocity pattern provides an estimation of left atrial pressure irrespective of the presence of left ventricular systolic dysfunction and, if not, to clarify the mechanism. BACKGROUND The pulsed Doppler transmitral flow velocity pattern, particularly peak early diastolic filling velocity, has been shown to change in parallel with left atrial pressure. However, extremely elevated left atrial pressure in association with heart failure does not necessarily cause an increase in peak early diastolic filling velocity in patients. METHODS Left atrial pressure was elevated with intravenous saline infusion in 11 dogs (normal left ventricular function group) and hemodynamic, transesophageal Doppler echocardiographic and M-mode echocardiographic variables were recorded at three different loading levels. In another 12 dogs, left atrial pressure was elevated by production of left ventricular systolic dysfunction with the stepwise injection of microspheres into the left coronary artery (left ventricular dysfunction group) and the same set of recordings was obtained at three different levels of dysfunction. RESULTS Peak early diastolic filling velocity increased with left atrial pressure in the normal left ventricular function group and correlated with mean left atrial pressure (r = 0.61, p < 0.01) and early diastolic left atrial to left ventricular crossover pressure (r = 0.71, p < 0.01). In contrast, peak early diastolic filling velocity did not increase with left atrial pressure in the left ventricular dysfunction group and did not correlate with mean left atrial pressure (r = -0.05) or the crossover pressure (r = 0.06). Peak early diastolic filling velocity correlated well with the difference between the crossover pressure and left ventricular minimal pressure in the left ventricular dysfunction group (r = 0.64, p < 0.01). In contrast to peak early diastolic filling velocity, deceleration time of the early diastolic filling wave correlated with mean left atrial pressure and the crossover pressure irrespective of the primary cause of preload alteration (r = -0.54, r = -0.59, p < 0.01 respectively, n = 69 for all data). CONCLUSIONS Preload dependency of the Doppler transmitral flow velocity pattern is hampered if an increase in left atrial pressure is due to left ventricular systolic dysfunction. In this setting, the increase in left ventricular minimal pressure due to left ventricular systolic dysfunction cancels the effect of the increase in left atrial pressure on the flow velocity pattern.