Ventilatory response after 1 min of voluntary hyperventilation (HV) was studied in 10 healthy women. Before, during, and after HV, end-tidal PCO2 (PETCO2) was maintained at a given level between resting and 60 Torr. After cessation of HV, hyperpnea was seen in 179 out of a total of 195 runs but in the remaining 16 runs in 3 subjects hypopnea occurred, both ventilatory changes being followed by slow recovery to the pre-HV level. The time constant (tau) of the decay process of post-HV hyperpnea was calculated and compared between the follicular (F) and luteal (L) phases of menstruation. For post-HV hypopnea, tau was assumed to be zero. There was an inverse correlation between tau and PETCO2 during the test, the relation being similar in F and L. With a phase change from F to L, tau value at resting PETCO2 increased from 17.7 to 23.7 s. Resting PETCO2 decreased from 40.8 to 37.7 Torr, and minute ventilation (VE) increased by 10%. The increased tau in L was ascribable to the decrease in resting PETCO2 but not to the increased ventilatory activity during the pre-HV period (corresponding to the resting VE) that was probably produced by ventilatory stimulation with progesterone in L. From these results, it is inferred that the ventilatory influence of progesterone might not be exerted on the brain stem, which has been implicated as a locus of the afterdischarge mechanism.