We previously demonstrated that, in awake goats, 6 h of hypoxic carotid body perfusion during systemic normoxia produced time-dependent hyperventilation that is typical of ventilatory acclimatization to hypoxia (VAH). The hypocapnic alkalosis that occurred could have produced VAH by inducing cerebral vasoconstriction and brain lactic acidosis even though systemic arterial normoxia was maintained. In the present study we tested the hypothesis that hypocapnic alkalosis is a necessary component of VAH. Goats were prepared so that one carotid body could be perfused, from an extracorporeal circuit, with blood in which gas tensions could be controlled independently from the blood perfusing the systemic arterial system, including the brain. Using this preparation we carried out 4 h of hypoxic carotid body perfusion while maintaining systemic arterial (and brain) normoxia in awake goats. Expired minute ventilation (VE) was measured while CO2 was added to inspired air to maintain normocapnia. Carotid body PCO2 and PO2 were maintained near 40 Torr during the 4-h carotid body perfusion. Control mean VE was 8.65 +/- 0.48 l/min (mean +/- SE). With acute carotid body hypoxia (30 min) VE increased to 21.73 +/- 2.02 l/min (P less than 0.05); over the ensuing 3.5 h of carotid body hypoxia, VE progressively increased to 39.14 +/- 4.14 l/min (P less than 0.05). These data indicate that neither cerebral hypoxia nor hypocapnic alkalosis are required to produce VAH. After termination of the 4-h carotid body stimulation, hyperventilation was not maintained in these studies, i.e., there was no deacclimatization. This suggests that acclimatization and deacclimatization are produced by different mechanisms.