The present study was undertaken to assess whether repeated measurement of cardiac 123I-metaiodobenzylguanidine (MIBG) imaging parameters before and after optimized treatments is useful for predicting the prognosis of patients with congestive heart failure (CHF) resulting from dilated cardiomyopathy (DCM). The subjects were 85 consecutive patients with DCM who had a left ventricular ejection fraction (LVEF) of less than 45%. The MIBG and the concentrations of neurohumoral factors were measured at baseline and after 6 months of optimized treatments. Cox proportional hazards analysis was performed to assess the various parameters before and after treatment. Twenty-three patients had a cardiac event (12 died; 11 hospitalized) during a mean follow-up period of 2 years. Although there was no difference between the baseline heart to mediastinum (H/M) ratio measured by MIBG between survivors and nonsurvivors, the H/M ratio was significantly decreased in nonsurvivors after 6 months. Multivariate analysis revealed that a high plasma concentration of brain natriuretic peptide level after 6 months (p=0.0049) and absolute changes in the H/M ratio (p=0.0046) were independent predictors of mortality. Comparison of the H/M ratio on MIBG imaging before and after optimized additional treatment provided useful information for predicting mortality and was independent of clinical and neurohumoral factors previously shown to be associated with poor prognosis in patients with DCM.