1. This study was designed to examine whether normal subjects could differentiate between the 'intensity' of their breathlessness and the amount of 'distress' it evoked, by specific wording of the instructions. 2. A preliminary study showed no significant difference between 'distress' score during exercise measured on two separate occasions (P = 0.3). 3. Ten subjects each performed two identical incremental cycle-ergometer exercise tests on separate occasions during which they were asked to quantify either 'intensity' or 'distress' by using modified Borg scales. 4. In all subjects there was a significant correlation (P less than 0.001) between 'intensity' and minute ventilation. In eight subjects there was a significant correlation (P less than 0.05) between 'distress' and minute ventilation. One subject displayed no significant correlation and one registered no distress. 5. Mean 'intensity' was greater than the mean 'distress' (P = 0.0001). The slope of 'intensity'/minute ventilation was greater than the slope of 'distress'/minute ventilation (P = 0.0001). 6. Within individuals there was a significant correlation between 'intensity' and 'distress' (P less than 0.05). There was a wide scatter in the slope of this relationship between subjects and maximum 'intensity' and 'distress' did not correlate. 7. Different elements of the breathlessness sensation could be identified and selectively measured depending on the wording of the instructions given to the subject. 8. There was a wide intersubject variation in the magnitude of both breathlessness 'intensity' and 'distress' estimates, but the differences between subjects in these two components of the sensation did not appear to follow a common pattern.