The aim of this study was to analyze the components of mitral and pulmonary A waves and to construct a Doppler-derived left ventricular (LV) end-diastolic pressure (EDP) prediction model based on the combined analysis of transmitral and pulmonary venous flow velocity curves. Combined analysis of transmitral and pulmonary venous flow velocity curves at atrial contraction is a reliable predictor of increased LV filling pressure. The duration of pulmonary and mitral A waves is determined by the sum of respective acceleration and deceleration time. Mitral flow and left upper pulmonary vein flow velocity curves were recorded simultaneously with LVEDP in 40 consecutive patients (aged 59 +/- 8 years) with coronary artery disease and preserved LV systolic function. Differences in all parameters represent values of pulmonary minus those of mitral A wave curve. The difference in deceleration time was the strongest candidate, being included in all models. After redundancy evaluation, we reached the following model: LVEDP = 20.61 + 0.229 x difference in deceleration time (r(2) = 0.80, p <0.001). In the entire study group, the difference in duration and in deceleration time of the A wave was highly correlated with LVEDP (r = 0.79, p <0.001, and r = 0.88, p <0.001, respectively). The entire study group was further divided according to whether LVEDP was above (group I, 20 patients) or below (group II, 20 patients) the median value (15.5 mm Hg). In group I, the difference in duration and in deceleration time correlated well (r = 0.62, p = 0.01, and r = 0.75, p = 0.001, respectively) with LVEDP, whereas in group II only the difference in deceleration time correlated well (r = 0.68, p = 0.005). In patients with coronary artery disease and preserved LV systolic function, the combined analysis of mitral and pulmonary A waves can predict LVEDP. The difference in deceleration time between pulmonary and mitral A waves can reliably evaluate high and normal LVEDP.