To examine the ability of myocardial contractile reserve (MCR) assessment to predict the improvement of left ventricular ejection fraction with treatment by carvedilol, a prospective study was undertaken in 85 patients with chronic heart failure and left ventricular ejection fraction < 45%. Low dose dobutamine echocardiography (DSE), a 6-min walk test and measured brain natriuretic peptide (BNP) were assessed in all the patients. Patients were separated into two groups. Group A were patients without any myocardial reserve and group B patients with a myocardial contractile reserve defined as an increment of more than 20% of the resting left ventricular ejection fraction during dobutamine infusion. The two groups differed for percentage of ischemic cardiomyopathy (67.8 in group A vs. 29.7% in group B P = 0.028), 6-min walk test performance (respectively, 343 vs. 415 meters P < 0.05) and BNP plasma levels (respectively, 184.5 vs. 70.1 P < 0.02) but not for left ventricular ejection fraction or NYHA class. During DSE, MCR and heart rate variation was higher in group B than in group A. At the end of the follow up, LVEF increased and NYHA class decreased in group B but not in group A. In multivariate analysis the existence of MCR could predict the improvement of LVEF with treatment by carvedilol. In our study, studying MCR could help to predict patients who will improve their LVEF with carvedilol prior to the administration of the treatment.