In order to determine whether or not resting ventilatory response to hypercapnia is changed by physical training, we studied the effect of long-term physical training on the slope of ventilatory response to CO2 at rest. The subjects were 9 untrained freshmen ranging in age from 18 to 20 years. Five out of nine subjects belonged to the badminton team after entering university in April 1980, and participated in their team's training for about 3 hr per day, 3 times a week year round for about 4 years until March 1984. Maximum oxygen uptake (VO2max), maximum pulmonary ventilation (VEmax) and maximum heart rate (HR max) were determined during maximal treadmill exercise before and after training. The slope (S) of ventilatory response to carbon dioxide at rest was measured by Read's rebreathing method. VO2max increased after training in the trained subjects and mean values of VO2max which were measured in 1982, 1983, and 1984, were statistically higher than that of 1980. Similar tendency was observed in VEmax and VO2max/W. Average values and standard deviations of S before training were 1.91 +/- 0.52 liter/min/torr and were decreased gradually with increasing training period; the differences in the S value before (1980) and after training, i.e., 1982, 1983, and 1984, were all significant. Such difference could still be seen after S was recalculated as SN by using normalized ventilation for 70 kg body weight, while there were no significant differences in the S and SN between baseline and repeated studies in the untrained group. In addition, CO2 responsiveness was found to correlate negatively with maximum oxygen uptake in 4 out of the 5 trained subjects. These results suggest that in normal subjects, long-term physical training, as in the present study, decreases CO2 responsiveness at rest.