Diminished ventilatory responsiveness to hypercapnia and hypoxia has been reported in athletes, but whether reduced chemosensitivity might lessen hyperpnea during exercise and facilitate performance is unknown. To evaluate ventilatory control and its possible role in athletic performance, we prospectively measured ventilatory responses to hypercapnia and hypoxia at rest and ventilatory equivalents for carbon dioxide (VE/VCO2) and oxygen (VE/VO2) during exercise in 20 accomplished marathon (42.2 km) runners (RUN) and 20 control subjects (CON). The athletes (mean age 27.8, range 18-41 yr) were all experienced runners with a mean best marathon time of 2 h, 36 min (range 2:14-2:55). There were no significant differences in the ventilatory responses at rest to hypercapnia (RUN, 2.23 +/- 0.73 vs. CON, 2.61 +/- 1.05 l X min-1 X Torr-1) and hypoxia (RUN, 0.57 +/- 0.40 vs. CON, 0.88 +/- 0.72 l X min-1 X 1% desat-1) (mean +/- SD). Similarly, there were no significant differences in VE/VCO2 and VE/VO2 between the two groups. Good correlation (r = 0.68; P less than 0.01) was observed between hypercapnic response at rest and exercise ventilation (VE/VCO2) in RUN. However, both hypercapnic and hypoxic ventilatory responses correlated poorly with marathon running time. These results demonstrate no differences in ventilatory responses at rest and during exercise between marathon runners and control subjects. The range of ventilatory responsiveness observed in this group of marathon runners indicates that a spectrum of ventilatory control is present in well-trained endurance athletes.