To evaluate whether the intact hypoxic drive from the carotid bodies during sustained hypoxia is required for the generation of hypoxic depression of ventilation (VE), 16 volunteers were exposed to two consecutive periods of isocapnic hypoxia (first period 20 min; second period 5 min; end-tidal PO2 45 Torr) separated by 6 min of normoxia. In study A, saline was given. In study B, 3 micrograms.kg-1.min-1 i.v. dopamine (DA), a carotid body inhibitor, was given during the first hypoxic exposure followed by saline during normoxia and the second hypoxic exposure. In study C, 20 min of normoxia with DA preceded 6 min of normoxia and 5 min of hypoxia without DA. The first peak hypoxic VE (PHV) in study A was approximately 100% above normoxic VE. After 20 min of hypoxia, VE declined to 60% above normoxic VE. The second PHV in study A was only 60% of the first PHV. We relate this delayed recovery from hypoxia to "ongoing" effects of hypoxic depression. During DA infusion, the changes in VE due to sustained hypoxia were insignificant (study B). The second PHV in study B was not different from the PHV after air breathing in studies A and C. This indicates that the recovery from sustained hypoxia with a suppressed carotid body drive was complete within 6 min. Our results show that despite central hypoxia the absence of ventilatory changes during 20 min of isocapnic hypoxia due to intravenous DA prevented the generation of central hypoxic depression and the depression of a subsequent hypoxic response.