The buffering capacity (beta) of rainbow trout (Oncorhynchus mykiss) plasma was manipulated prior to intravascular injection of bovine carbonic anhydrase to test the idea that proton (H+) availability limits the catalysed dehydration of HCO3- within the extracellular compartment. An extracorporeal blood shunt was employed to continuously monitor blood gases in vivo in fish exhibiting normal plasma beta (-3.9+/-0.3 mmol 1(-1) pH unit(-1)), and in fish with experimentally (using N-[2-hydroxyethyl]piperazine-N'-[2-ethanesulfonic acid]) elevated plasma beta (-12.1+/-1.1 mmol 1(-1) pH unit(-1)). An injection of 5 mg kg(-1) carbonic anhydrase equally reduced (after 90 min) the arterial partial pressure of CO2 in trout with regular (-0.23+/-0.05 Torr) or high (-0.20+/-0.05 Torr) plasma beta; saline injection was without effect. Because ventilation and venous blood gases were unaffected by carbonic anhydrase, the effect of extracellular carbonic anhydrase in lowering arterial partial pressure of CO2 was likely caused solely by a specific enhancement of CO2 excretion owing to acceleration of HCO3- dehydration within the plasma. The lowering of arterial partial pressure of CO2 in trout after injection of exogenous carbonic anhydrase provides the first in vivo evidence that the accessibility of plasma HCO3- to red blood cell carbonic anhydrase constrains CO2 excretion under resting conditions. Because the velocity of red blood cell Cl-/HCO3- exchange governs HCO3- accessibility to red blood cell carbonic anhydrase, the present study also provides evidence that CO2 excretion at rest is limited by the relatively slow rate of Cl-/HCO3- exchange. The effect of carbonic anhydrase in lowering arterial partial pressure of CO2 was unrelated to plasma buffering capacity. While these data could suggest that H+ availability does not limit extracellular HCO3- dehydration in vivo at resting rates of CO2 excretion, it is more likely that the degree to which plasma beta was elevated in the present study was insufficient to drive a substantially increased component of HCO3- dehydration through the plasma.