To determine the relationship of ventilatory responsiveness to hypoxia and hypercapnia to exercise hyperpnea, these responses and steady-state exercise ventilation (VE) were measured in 16 athletes during light (1/3 VO2 max) and heavy (2/3 VO2 max) exercise. Both the hypoxic and hypercapnic ventilatory responses correlated positively with VE per unit metabolic rate (VE/VCO2) at both exercise levels (P less than 0.05). The contribution of the hypoxic response to normoxic exercise VE was quantified by comparing VE in normoxia to VE during a brief (1 min) exposure to high O2 (PAO2 = 200 Torr). High O2 reduced normoxic exercise VE by a mean of 20% at either exercise intensity. Among individuals this reduction was directly dependent upon the intensity of the hypoxic response, and ranged from 7 to 42% of normoxic VE. After the variable reduction of normoxic VE by hyperoxia, all correlations of ventilatory response with exercise VE were lost except for the correlation of hypercapnic response with heavy exercise VE/VCO2. These findings indicate that the extent of VE in light or heavy exercise is modified by the strength of the hypoxic ventilatory response, and that the hypercapnic response independently correlates with VE during heavy exercise.