The purpose of this study was to determine whether the increased ventilation/perfusion (VA/Q) mismatch caused by hypocapnic hyperventilation in dogs (J. Appl. Physiol. 1993; 74:1306-1314) is a direct CO2 or a pH-mediated effect. From an initial state of hyperventilated respiratory alkalosis (FIO2 = 0.21, VT = 18 ml/kg, RR = 35), we studied the changes in VA/Q distributions, respiratory gas exchange, and hemodynamics when the acid-base status of the dogs was manipulated by combinations of acid or alkali infusion with or without CO2 inhalation. In this manner, we studied respiratory alkalosis (high pH, low PCO2), normalized acid-base status (normal pH, normal PCO2), metabolic acidosis (low pH, normal PCO2), metabolic alkalosis (high pH, normal PCO2), and a mixed respiratory alkalosis and metabolic acidosis (normal pH, low PCO2). Gas exchange was evaluated using the multiple inert gas elimination technique. PaO2 was reduced and VA/Q heterogeneity was increased in all conditions defined by a high pH, independent of the PCO2 (respiratory alkalosis and metabolic alkalosis). In contrast, PaO2 and VA/Q heterogeneity was unchanged in conditions defined by either a normal or low pH (normalized acid-base status, mixed respiratory alkalosis and metabolic acidosis, and metabolic acidosis). Therefore, we conclude that hypocapnia-induced VA/Q mismatch in hyperventilated dogs is pH-mediated and is not a function of PCO2 per se.