Abstract Arterial chemoreceptor discharge and ventilation are both significantly increased when the concentration of arterial potassium is raisedto a level typical of moderate exercise. However, although the plasma potassium level of exercising, beta-blocked patients rises by more than that of normal subjects, this does not show up in their steady-state ventilatory response, i.e. exercising beta-blocked subjects ventilate no more than exercising controls. The present experiments were designed to test the hypothesis that the apparent failure of beta-blocked subjects to respond to the extra hyperkalaemia that they experience might be accounted for by a reduction in the sensitivity of arterial chemoreceptors to potassium. We used eleven pentobarbitone-anaesthetized, thoracotomized, artificially hyperventilated cats, in which arterial potassium was raised from ca. 4.5 to ca. 7 mM before and during beta blockade by propranolol or atenolol. The steady-state relation between chemoreceptor discharge and arterial potassium was curvilinear, discharge becoming more sensitive to potassium as the concentration of the latter was raised. Beta blockade significantly reduced discharge at all levels of plasma potassium ( P<0.0001). It also significantly reduced ( P<0.05) the slope of the response of discharge to a given increase of plasma potassium. Our results show that beta blockade decreases the sensitivity of arterial chemoreceptors to increases in arterial potassium. This may explain why exercising beta-blocked subjects breathe no harder than controls, in spite of the fact that they are more hyperkalaemic.