The upper limit of incidence of muscle sympathetic neural bursts can lead to underestimation of sympathetic activity in patients with severe heart failure. This study aimed to evaluate the pulse-synchronous burst power of muscle sympathetic nerve activity (MSNA) as a more specific indicator that could discriminate sympathetic activity in patients with heart failure. In 54 patients with heart failure, the pulse-synchronous burst power at the mean heart rate was quantified by spectral analysis of MSNA. Thirteen patients received a central sympatholytic agent (guanfacine) for 5 days to validate the feasibility of this new index. Both burst incidence and plasma norepinephrine level showed no significant difference between patients in New York Heart Association functional class III (94 +/- 6 per 100 heartbeats and 477 +/- 219 pg/ml, respectively) and class II (79 +/- 14 per 100 heartbeats and 424 +/- 268 pg/ml, respectively). In contrast, the burst power was useful for discriminating patients in class III from those in class II (61 +/- 8% vs. 39 +/- 10%; P < 0.05). Inhibition of sympathetic nerve activity by guanfacine was more sensitively reflected by the change of burst power (-36 +/- 25%) than by that of burst incidence (-12 +/- 14%; P < 0.001). The sympathetic burst power reflects both burst frequency and amplitude independently of the absolute values and provides a sensitive new index for interindividual comparisons of sympathetic activity in patients with heart failure.