Cardiac resynchronization therapy (CRT) decreases muscle sympathetic nerve activity (MSNA) in patients with severe congestive heart failure (CHF) and cardiac asynchrony. Whether this affects equally patients who clinically respond or not to CRT is unknown. We tested the hypothesis that the favorable effects of CRT on MSNA disappear on CRT interruption only in those who respond to CRT. Twenty-three consecutive CHF patients participated in the study, among whom 16 presented a symptomatic improvement by one or more New York Heart Association (NYHA) functional classes 15 +/- 5 mo after CRT (responders), and seven had not improved after 12 +/- 4 mo of CRT (nonresponders). MSNA and echocardiographic recordings were obtained in random order during atrio-right ventricular pacing (ARV), without stimulation in patients who were not pacemaker dependent (OFF, n = 17), and during atrio-biventricular pacing (BIV). Responders had a longer 6-min walking distance, a lower NYHA class and brain natriuretic peptide levels, and a better quality of life than did nonresponders (all P < 0.05). MSNA increased by 25 +/- 7% in the responders, whereas it remained unchanged in the nonresponders, when shifting from the BIV mode to a nonsynchronous condition (ARV and OFF modes) (P < 0.01). Cardiac output decreased by 0.7 +/- 0.2 l/min in the responders but did not change when shifting from the BIV mode to the nonsynchronous pacing mode in the nonresponders (P < 0.01). In conclusion, reversible sympathoinhibition is a marker of the clinical response to CRT.