Influences of slow and deep respiration on steady-state sympathetic nerve activity remain controversial in humans and could vary depending on disease conditions and basal sympathetic nerve activity. To elucidate the respiratory modulation of steady-state sympathetic nerve activity, we modeled the dynamic nature of the relationship between lung inflation and muscle sympathetic nerve activity (MSNA) in 11 heart failure patients with exaggerated sympathetic outflow at rest. An autoregressive exogenous input model was utilized to simulate entire responses of MSNA to variable respiratory patterns. In another 18 patients, we determined the influence of increasing tidal volume and slowing respiratory frequency on MSNA; 10 patients underwent a 15-min device-guided slow respiration and the remaining 8 had no respiratory modification. The model predicted that a 1-liter, step increase of lung volume decreased MSNA dynamically; its nadir (-33 +/- 22%) occurred at 2.4 s; and steady-state decrease (-15 +/- 5%), at 6 s. Actually, in patients with the device-guided slow and deep respiration, respiratory frequency effectively fell from 16.4 +/- 3.9 to 6.7 +/- 2.8/min (P < 0.0001) with a concomitant increase in tidal volume from 499 +/- 206 to 1,177 +/- 497 ml (P < 0.001). Consequently, steady-state MSNA was decreased by 31% (P < 0.005). In patients without respiratory modulation, there were no significant changes in respiratory frequency, tidal volume, and steady-state MSNA. Thus slow and deep respiration suppresses steady-state sympathetic nerve activity in patients with high levels of resting sympathetic tone as in heart failure.