The intensity of breathlessness induced by ventilatory stimulation resulting from hypercapnia, hypoxia or exercise has been quantified in normals by using the two different sensory scaling techniques of linear visual analogue scaling and ratio magnitude estimation. In naive individuals both techniques show good face validity. When related to ventilation, quantification of breathlessness is moderately reproducible with both methods, even when subjects are kept in ignorance of the pattern of ventilatory stimulation. There is a small within- and large between-subject variability with both scaling techniques; possible factors responsible are discussed. The reproducibility of visual analogue scaling when related to ventilation is independent of the nature of the ventilatory stimulus and is maintained over intervals as long as 1 week when memory for the score given is unlikely to be an important factor. The difficulties of interpreting subjective estimates of perceived breathlessness are discussed, together with the relative merits of the two scaling techniques.