We tested the hypothesis that, in humans, hyperthermic hyperpnea elicited in resting subjects differs from that elicited during submaximal, moderate-intensity exercise. In the rest trial, hot-water legs-only immersion and a water-perfused suit were used to increase esophageal temperature (T(es)) in 19 healthy male subjects; in the exercise trial, T(es) was increased by prolonged submaximal cycling [50% peak O(2) uptake (Vo(2))] in the heat (35 degrees C). Minute ventilation (Ve), ventilatory equivalent for Vo(2) (Ve/Vo(2)) and CO(2) output (Ve/Vco(2)), tidal volume (Vt), and respiratory frequency (f) were plotted as functions of T(es). In the exercise trial, Ve increased linearly with increases (from 37.0 to 38.7 degrees C) in T(es) in all subjects; in the rest trial, 14 of the 19 subjects showed a T(es) threshold for hyperpnea (37.8 +/- 0.5 degrees C). Above the threshold for hyperpnea, the slope of the regression line relating Ve and T(es) was significantly greater for the rest than the exercise trial. Moreover, the slopes of the regression lines relating Ve/Vo(2), Ve/Vco(2), and T(es) were significantly greater for the rest than the exercise trial. The increase in Ve reflected increases in Vt and f in the rest trial, but only f in the exercise trial, after an initial increase in ventilation due to Vt. Finally, the slope of the regression line relating T(es) and Vt or f was significantly greater for the rest than the exercise trial. These findings indicate that hyperthermic hyperpnea does indeed differ, depending on whether one is at rest or exercising at submaximal, moderate intensity.