Determining response dynamics of hypoxic air hunger may provide information of use in clinical practice and will improve understanding of basic dyspnea mechanisms. It is hypothesized that air hunger arises from projection of reflex brain stem ventilatory drive ("corollary discharge") to forebrain centers. If perceptual response dynamics are unmodified by events between brain stem and cortical awareness, this hypothesis predicts that air hunger will exactly track ventilatory response. Thus, during sustained hypoxia, initial increase in air hunger would be followed by a progressive decline reflecting biphasic reflex ventilatory drive. To test this prediction, we applied a sharp-onset 20-min step of normocapnic hypoxia and compared dynamic response characteristics of air hunger with that of ventilation in 10 healthy subjects. Air hunger was measured during mechanical ventilation (minute ventilation = 9 +/- 1.4 l/min; end-tidal Pco(2) = 37 +/- 2 Torr; end-tidal Po(2) = 45 +/- 7 Torr); ventilatory response was measured during separate free-breathing trials in the same subjects. Discomfort caused by "urge to breathe" was rated every 30 s on a visual analog scale. Both ventilatory and air hunger responses were modeled as delayed double exponentials corresponding to a simple linear first-order response but with a separate first-order adaptation. These models provided adequate fits to both ventilatory and air hunger data (r(2) = 0.88 and 0.66). Mean time constant and time-to-peak response for the average perceptual response (0.36 min(-1) and 3.3 min, respectively) closely matched corresponding values for the average ventilatory response (0.39 min(-1) and 3.1 min). Air hunger response to sustained hypoxia tracked ventilatory drive with a delay of approximately 30 s. Our data provide further support for the corollary discharge hypothesis for air hunger.