In a magnitude production test, subjects of both sexes rated their bre
ath depth under different conditions. In the first group, 50 subjects
were examined at rest in the supine position and breathing spontaneous
ly or assisted by a respirator. The second group, consisting of 33 sub
jects, was tested in the sitting position at rest and during mild bicy
cle exercise (females 35 W, males 40 W). Subjectively quantified breat
h depths and corresponding tidal volumes were compared and their relat
ions were described by Stevens power function. The assessment of breat
h depth in spontaneous breathing at rest was remarkably precise (mean
Stevens exponent 0.94+/-0.02). During assisted ventilation, rating was
less accurate and tended to underestimate breath depth. Noticeable un
der both conditions was a tendency to overestimate voluntary tidal vol
ume reductions. This was more pronounced in young (up to 35 years) wom
en than in men or elderly women. During exercise, the subjects (men an
d women) overestimated both increases and reductions of tidal volume.
Each voluntary tidal volume modification was accompanied immediately b
y involuntary changes of inspiration and expiration times, thus, parti
ally compensating disturbances of ventilation. In addition, voluntaril
y reduced tidal volumes were quantitatively compensated within the fir
st breath subsequent to the voluntary manoeuver. We conclude that brea
th depth sensation is more strongly related to proprioceptive signals
than to visceroceptive signals and is influenced by the autonomous res
piratory drive. Furthermore, proprioception is assumed to participate
in the control of the autonomous respiratory drive.