The aim of this study was to assess muscle sympathetic nerve activity (MSNA) when the work of breathing is experimentally lowered during dynamic exercise. Healthy subjects (n=12; age=29±9years) performed semi-recumbent cycling trials at 40%, 60%, and 80% of peak workload. Exercise trials consisted of spontaneous breathing, reduced work of breathing (proportional assist ventilator), followed by further spontaneous breathing (post-ventilator). MSNA was recorded from the median nerve. There was no difference in work of breathing between PAV and post-PAV at 40% peak work. At 60% peak work, the ventilator significantly (P<0.05) reduced work of breathing (103±39 vs 144±47Jmin-1 ), sympathetic nerve activity (35±5 vs 42±8burstmin-1 ), and (2.4±0.5 vs 2.6±0.5Lmin-1 ) without influencing ventilation (86±9 vs 82±10Lmin-1 ; P>0.05), for PAV and post-PAV respectively. During 80% peak work (n=8), the ventilator significantly (P<0.05) reduced work of breathing (235±110 vs. 361±150Jmin-1 ), MSNA (48±7 vs 54±11burstmin-1 ), and (2.9±0.6 vs 3.2±0.7Lmin-1 ) but not ventilation (121±20 vs 123±20Lmin-1 ; P>0.05), for PAV and post-PAV respectively. There was a significant relationship between MSNA and (P<0.0001) with a significant interaction due to the ventilator (P<0.05). Lowering the normally occurring work of breathing during exercise results in commensurate reductions in MSNA. Our findings provide evidence of a sympathetically mediated vasoconstrictor effect emanating from respiratory muscles during exercise.