The transient changes in ventilation and alveolar PCO2 following the inhalation of 4% CO2 and the return to air have been studied in five normal subjects in an attempt to characterize the intact respiratory chemostat in man. In all 13 studies the changes of ventilation lagged behind those of alveolar PCO2, while the latter consistently over- or undershot its later steady-state value. The timing and magnitude of the responses were in general agreement with the predictions made by Grodins et al., who assumed in their model that the chemoreceptors respond to a single stimulus (PCO2) and are embedded in tissue in equilibrium with venous blood. In four duplicate experiments on individual subjects the quantitative reproducibility of the transients was poor, due largely to cyclic and random variations of ventilation shown by all subjects, even when in the steady state. This precluded any accurate search for differences introduced into the predicted transient response by the adoption of Defares' more sophisticated model, in which the influence of a changing cerebral blood flow on receptor PCO2 is considered. Studies performed during mild hypoxia revealed no change in transient response. control of ventilation; steady-state ventilation; random changes in ventilation; cyclic changes in ventilation; effect of hypoxia on ventilation; respiratory chemoreceptor Submitted on June 24, 1963