To determine the prognostic significance of pulsed wave Doppler-derived left ventricular diastolic filling velocity profiles and the relationship between Doppler variables and clinial functional status, the follow-up outcomes of 62 consecutive patients with dilated cardiomyopathy and symptoms of left ventricular dysfunction were analyzed. All patients had echocardiographic left ventricular end-diastolic dimension ≥ 6.0 cm, fractional shortening < 25%, increased E pointseptal separation, and diffuse hypokinesia or akinesia. During a mean follow-up period of 30.5 ± 13.9 months, 27 patients experienced cardiac events: 23 died of either progressive pump failure or an episode of sudden death and four required cardiac transplantation because of refractory heart failure. Peak early filling velocity (78 ± 23 cm/sec vs 65 ± 25 cm/sec; p < 0.03) was higher and late atrial filing velocity (34 ± 13 cm/sec vs 55 ± 19 cm/sec; p < 0.001) was lower in patients with cardiac events than in cardiac event-free survivors. The ratio of early to late transmitral filling velocities was higher (2.6 ± 1.2 vs 1.5 ± 1.3; p < 0.001), and the deceleration time of early diastole was shorter (133 ± 48 msec vs 175 ± 71 msec; p < 0.001) in patients with cardiac events. The cardiac event rate was significantly higher in patients with an early to late filling velocity ratio greater than 2 (77% vs 19%; p < 0.001) or a deceleration time less than 150 msec (58% vs 23%; p < 0.05) than in those without. Stepwise multivariate regression analysis revealed that the pattern of transmitral early to late filling velocity ratio was the only significant independent Doppler echocardiographic predictor of outcome for these patients. Repeat Doppler echocardiographic examinations, which were performed in 31 survivors after intensive treatment (mean, 38.6 ± 6.5 months), showed that early filling velocity was decreased (55 ± 20 cm/sec vs 75 ± 25 cm/sec; p < 0.02), late atrial filling velocity was increased (74 ± 27 cm/sec vs 57 ± 21 cm/sec; p < 0.01), early to late filling velocity ratio was reduced (0.8 ± 0.3 vs 1.7 ± 1.3; p < 0.001), and deceleration time was prolonged (227 ± 60 msec vs 167 ± 82 msec; p < 0.01) in 18 patients with clinical functional improvement, whereas these measurements were unaltered in the remaining 13 patients whose functional status was unchanged or had deteriorated. The occurrence of either an early to late filling velocity ratio greater than 2 or a deceleration time less than 150 msec was significantly more frequent among in patients without functional improvement. The study suggests that pulsed wave Doppler-derived left ventricular diastolic filling variables may be important predictors of outcome in patients with dilated cardiomyopathy and may provide useful measures in determining the effects of therapy during long-term follow-up of these patients.