We used a helium-rebreathe technique in nine healthy subjects to determine the effects of exercise intensity and duration on end-expiratory lung volume (EELV). The rebreathe functional residual capacity (FRC) technique was shown: (a) to be similar to that measured in the body plethysmograph, at rest; (b) to agree closely with volitionally induced changes in EELV as determined by inductance plethysmography, at rest; (c) to be reproducible within subjects between trials conducted at rest or exercise on different days (r = 0.96, coefficient of variation +/- 3%); (d) to correlate significantly with coincident changes in end-expiratory esophageal pressure from rest to exercise, with increasing exercise intensity and over time at a constant exercise load. Exercise-induced reductions in EELV occurred in all subjects, averaging 0.3 L (-0.1 to -0.7 L) in light exercise and 0.79 L (-0.5 to -1.2 L) in heavy or maximum exercise. This reduction in EELV accounted for slightly more than one-half of the increase in VT during light exercise and slightly less than one-half of the increased VT in heavy exercise. In heavy prolonged exercise lasting 8-15 min, EELV fell in the initial 2 min and was either sustained at this reduced level or fell further with exercise duration to exhaustion. We found that FRC was reduced even in very light exercise when changes in TE and VE from rest were minimal; further reductions in EELV occurred as end-inspiratory lung volume increased and expiratory time shortened with increasing exercise intensity and duration. Based on these types of changes we speculate that active expiration during exercise in humans may be controlled by a combination of locomotor-related feed-forward and lung volume related feed-back mechanisms.