The present study was performed to evaluate how to assess cardiac resynchronization therapy (CRT) efficacy in chronic heart failure (CHF) through clinical, echocardiographic and exercise analysis. We analyzed 41 stable CHF (NYHA III) patients with: left bundle-branch-block, ejection fraction <35%, left-ventricular dissynchrony (by tissue-Doppler), peak oxygen consumption (VO2) <16 ml/kg/min, suitable cardiac vein (by multislice computed tomography) and no anemia or kidney failure. Patients were evaluated before and after (7+/-3 months) CRT. Two patients died. CRT responders to none of the evaluated criteria were 19.5%. The best agreement (90%) with clinical response was obtained using the presence/absence of either left-ventricular systolic volume (LVSV) or peakVO2 response. In less severe CHF (peakVO2 12-16 ml/kg/min), peakVO2 and work-load didn't change after CRT, despite echocardiographic, ventilation/carbon dioxide relationship and clinical improvement. Echocardiography and CPET are complementary for the evaluation of CRT, but not in less severe CHF patients, where the role of CPET remain uncertain.