Rationale: Central processing of dyspnea relief remains largely unknown. Objectives: To identify physiological determinants, quality of sensation and brain activation associated with dyspnea relief. Measurements and Main Results: Dyspnea relief was induced in 10 healthy volunteers by decreasing an adjustable external resistive load (~ 15 - 50 cm H₂O/L/s). Brain imaging (positron emission tomography) was performed during either dyspnea or relief. Perceived intensity of moderate and high relief was similar to that of its preceding dyspnea (Borg scores = 5.10 ± 1.49 versus 5.3 ± 1.4, and 2.78 ± 0.94 versus 2.99 ± 0.94, respectively, P < 0.05) and was predominantly related to reversal of dyspnea-induced increased mouth pressure/ventilation ratio (r² = 0.88, P < 0.001). Dyspnea relief involved specific, mostly positivevalenced descriptors i.e. breathing-related pleasure and/or reward. Most significant reliefassociated brain activation was detected in the left anterior cingulate cortex (Z = 4.7, corrected P < 0.05) and additional activation (uncorrected P < 0.0001) in the posterior cerebellum and in the temporal and prefrontal cortices. For dyspnea, significant activation was located in the right caudate nucleus, the anterior cerebellum (Z = 5 and 4.65, respectively, corrected P < 0.05) and the premotor cortex, while deactivation occurred in the left prefrontal cortex (Z = 4.11). Conclusions: Relief of acutely load-induced dyspnea is not simply the neutral perception of dyspnea decrease but rather a strong, positive-valenced sensation that is associated with characteristic brain activation distinct from that subserving dyspnea perception and possibly reflecting activation of a dyspnea modulation network.