To test the passive transport hypothesis of cerebrospinal fluid (CSF) [H+] regulation, we altered the relationship between plasma [H+] and the electrical potential difference between CSF and blood (PD) by elevating plasma [K+] during 6-h systemic acid-base disturbances. In five groups of pentobarbital-anesthetized dogs, we increased plasma [K+] from 3.5 to an average of 7.8 meq/l. Hyperkalemia produced an increase in the PD of 6.3 mV by 6 h with normal plasma acid-base status (pHa 7.4), of 8.3 mV with isocapnic metabolic acidosis (pHa 7.2), of 5.3 mV with isocapnic metabolic alkalosis (pHa 7.6), of 9.2 mV with isobicarbonate respiratory acidosis (PaCO2 61 Torr) and of 5.7 mV with isobicarbonate respiratory alkalosis (PaCO2 25 Torr). The change in CSF [H+] at 6 h in each group was the same as that observed in normokalemic animals (Am. J. Physiol. 228: 1134-1154, 1975). This result is not consistent with the passive transport hypothesis. The CSF-blood PD is therefore not an important determinant of CSF [H+] CSF [H+] homeostasis must result from some form of active transport control.