ABSTRACT A transient inhibitory effect of catecholamines on relative CO2 excretion, mediated by an inhibition of HCO3− dehydration through the red blood cell (RBC), has been proposed to cause the increase in routinely observed after strenuous exercise in fish (‘CO2 retention hypothesis’, Wood and Perry, 1985). To evaluate this idea, trout fitted with arterial cannulae, oral membranes and opercular catheters were placed in ventilation chambers. , RBC intracellular pH (pHi) and other blood acid-base parameters were monitored from the arterial cannulae. The ventilation chamber system allowed continuous, almost instantaneous, measurements of water ΔO2 and ΔCO2 across the gills, and therefore of respiratory exchange ratio (RE), as well as Δammonia, mean expired pH and ventilation volume . Physiological doses of adrenaline and noradrenaline (3.2 nmol kg−1), designed to duplicate typical post-exercise concentrations, together with appropriate saline controls, were injected into resting fish. Adrenaline caused an immediate hypoventilation, while the response to noradrenaline was biphasic: hyperventilation followed by hypoventilation. With both drugs, ΔO2 and ΔCO2 increased, but RE remained constant (adrenaline) or increased (noradrenaline). There was no evidence of a specific inhibition of CO2 excretion, nor was there any increase in ; changes in RBC pHi were small (noradrenaline) or non-existent (adrenaline). These results confirm those of Steffensen el al. (1987) and do not support the CO2 retention hypothesis. However, the RBCs of resting trout may be relatively insensitive to catecholamines at normal arterial blood pH (pHa).