The respiratory distress of breathholding has been shown to be relieved by breathing, even without correctino of worsening blood gases (Fowler, 1954). We repeated the study by having untrained normal subjects perform maximal breathholds which were followed by the rebreathing of a gas mixture containing 7.5% CO2 and 8.2% O2, and then by second breathholds. In addition, we had the subjects continuously rate thei respiratory distress using a visual analog scale (VAS). The rating were easy to perform and were highly reproducible on repeated trials in a given subject. Subjects experienced increasing distress during the breathhold, rapid and substantial relief upon rebreathing, and then were capable of performing second breathholds, all consistent with Fowler's results. The findings are consistent with animal studies in which a neural mechanism associated with stimulation of pulmonary stretch receptors inhibits the firing of midbrain neurons which may be involved in transmission to the cortex of sensory information about breathing. © 1994.