To investigate the effect of the neural components associated with leg movements on the control of ventilation during recovery from exercise, we recorded the minute ventilation (VE), oxygen uptake (VO2), and carbon dioxide output (VCO2) of eight normal volunteers during recovery from moderate, steady-state cycle exercise (170 W). The recovery phases were undergone separately under two different conditions: 5 min of rest (passive recovery) on a bicycle ergometer and 3 min of pedaling at a work rate of 0W (active recovery) followed by 2 min of rest. The phase-1 responses were observed in all the variables studied at the transition of passive recovery but not in the active recovery phase. The kinetics of VCO2, during the off-transition were significantly faster than those of VE in both recoveries, indicating that the decreases in VCO2 could precede the decreases in VE. Although the levels of VE and VCO2 during active recovery were significantly higher than those during passive recovery, the decline in VE was closely proportional to that of VCO2 under both recovery conditions, with resultant indications of similar VE-VCO2 regression lines. These findings suggest that the flux of CO2 to the lungs is an important determinant of ventilatory drive during recovery, and that neither central command nor neural afferents from contracting muscles are requisite for the control of ventilation during recovery from exercise.