The ventilatory response to submaximal exercise, defined as the slope of minute ventilation over carbon dioxide production (VE/VCO2), was determined in 12 normal subjects, ten patients with pulmonary hypertension before and after heart-lung transplantation, and eight patients following heart transplantation. Patients with pulmonary hypertension show an augmented ventilatory response compared to normal subjects (pulmonary hypertension [mean, 57.7 +/- 6.8 (SE) ml/ml VCO2; normal subjects, 22.3 +/- 1.4 ml/ml VCO2; p less than 0.001]). Following heart-lung transplantation, VE/VCO2 slope fell to 24.7 +/- 1.6 ml/ml VCO2, a value which is not significantly different than the value in normal subjects. Patients after heart transplantation show a mean slope value of 25.3 +/- 1.3 ml/ml VCO2, which is not significantly different than the normal value or the value found after heart-lung transplantation. The augmented ventilatory response to exercise did not correlate with the usual chemical modulators of ventilation (arterial pH, arterial carbon dioxide tension, or arterial oxygen tension). These results suggest the following: the existence of a neural system in patients with pulmonary hypertension which results in an augmentation of ventilatory drive in response to exercise; the augmented ventilatory response reflects excessive neural activity of pulmonary afferents during exercise; narrow regulation of the ventilatory response to exercise in normal subjects which is preserved in the denervated lung, indicating that pulmonary afferents are not critical to ventilatory control during exercise in the normal subject; and the possible use of measurements of the ventilatory response to exercise as a noninvasive screening test for pulmonary hypertension.