In patients with chronic heart failure and reduced systolic function, an implantable cardioverter-defibrillator (ICD) improves the prognosis, but morbidity and mortality remain high. We attempted to identify the prognostic impact of Doppler echocardiography and QRS duration in such patients. We prospectively enrolled 84 patients with chronic heart failure, an ICD, and impaired systolic function (mean ejection fraction 29 +/- 10%). Echocardiographic measurements included left ventricular dimensions/volumes, ejection fraction, mitral E/A ratio, deceleration time, and tissue Doppler analysis of mitral annular velocities (S', E', A'). A cardiac event (death from pump failure or appropriate ICD therapy, i.e., antitachycardia pacing/shock due to sustained ventricular tachycardia or ventricular fibrillation) was defined as the study end point. During a follow-up of 373 +/- 254 days, 22 patients (26%) had an event (death from pump failure, n = 7; patients who received an appropriate ICD therapy, n = 16). In patients with an event, the QRS duration was longer (169 +/- 41 vs 146 +/- 37 ms, p = 0.023), the mitral E/E' ratio was higher (16.0 +/- 6.5 vs 12.8 +/- 5.9, p = 0.044), and a restrictive filling pattern was more frequent (44% vs 9%, p = 0.017). Stepwise multivariate Cox regression analysis identified a restrictive filling pattern as the only independent predictor of an event (hazard ratio 3.65, 95% confidence interval 1.54 to 8.64, p = 0.003). For patients with a restrictive filling pattern, the outcome was markedly poorer than that for patients with a nonrestrictive pattern (event-free survival rate 38% vs 72%, p = 0.005). In conclusion, in patients with chronic heart failure, an ICD, and systolic dysfunction, a restrictive filling pattern is an independent predictor of adverse cardiac events.