Background: The evaluation of dynamic intra-ventricular asynchrony has been reported to be feasible. But, its prognostic implication in functional improvement (FI) has not been investigated. Methods: Supine bicycle exercise was performed in 41 patients with non-ischemic cardiomyopathy (52짹12 years, EF<40%). Using tissue Doppler imaging, averages of peak systolic velocities of 6 basal segments were obtained at rest (V_b) and peak exercise (V_p) to assess the longitudinal contractility. Dyssynchrony index (Ts-SD12) was defined as the standard deviation of time to peak systolic velocities at 12 LV segments. The percentage change of end-systolic volume (ESV) during exercise was calculated as [(ESV_p- ESV_b)/ESV_bx 100 (%)]. At least a 15% reduction in ESV at follow-up (%��ESV_f <-15%) was considered as FI. Results: Intra-observer and inter-observer variability of Ts-SD12 were 8% and 9% measured at rest and 12% and 14% at peak exercise. During exercise for 8.7짹2.0 min, the mean value of Ts-SD12 was not changed (27짹12���27짹13 ms), whereas changes in dyssynchrony (��Ts-SD12_p) were individually variable. Although Ts-SD12_b and dynamic changes in MR correlated with Ts-SD12_p, V_p was an independent determinant for Ts-SD12_p (棺=-0.541, p<0.001). Follow-up echo was performed in 35 patients after appropriate medication for 16짹3 months. %��ESV_f showed significant correlation with changes in MR(r=0.434, p=0.009), disease duration (r=0.343, p=0.045), QRS interval (r=0.347, p=0.041) and ��Ts-SD12_p (r=0.511, p=0.002). However, multivariate analysis identified Ts-SD12_p as the strongest predictor for FI (%��ESV_f: 棺=0.577, p<0.001 and ��EF_f: 棺=-0.563, p<0.001). On ROC analysis, the sensitivity and specificity of Ts-SD12_p < 27 ms was 70% and 75% to predict %��ESV_f <-15% (AUC=0.72, 95% CI 0.55-0.89, p=0.027). Conclusion: In heart failure patients, the evaluation of dynamic dyssynchrony was feasible and Ts-SD12_p was the most powerful predictor for LV remodeling and response to medical treatment.