Our objective was to gain insight into the role of the carotid chemoreceptors (CC) in the exercise hyperpnea. Humans and ponies were studied at rest and during submaximal exercise breathing room air. In healthy humans, alveolar ventilation (V-A) was tightly matched to CO2 production (CO2) resulting in PaCO2 deviating during exercise less than 1-2 mm Hg from rest. In contrast, ponies' V-A increased proportionately more than VCO2 during exercise resulting in a workload dependent hypocapnia. Attenuating CC activity through hyperoxia had no effect on exercise PaCO2 of humans but hyperoxia accentuated the exercise hypocapnia of ponies. Similarly, CC denervation accentuated the exercise hypocapnia of ponies. Healthy humans were also studied while external airway resistance was increased, which, while breathing room air, resulted in a workload dependent hypercapnia, and this hypercapnia was accentuated by hyperoxia. Finally, a majority of asthmatic humans studied were hypercapnic during exercise while breathing room air and the hypercapnia was accentuated by hyperoxia. We conclude that the CC do not provide a primary drive for the exercise hyperpnea but they 'fine tune'' V-A to minimize disruptions of arterial blood gases. In healthy humans, attenuating CC activity has no effect on PaCO2 because the primary V-A drive is closely matched to VCO2.