The study of a patient with an Arnold Chiari malformation gave us the opportunity to test the hypothesis that the motor cortex contributes significantly to respiratory control during muscular exercise through the corticospinal pathway. The patient was a 25 years old woman who exhibited a severe impairment of the 'automatic' ventilatory control due to a type I Arnold Chiari malformation. Since she never complained of being breathlessness even on exertion, the breath-by-breath ventilatory (VE) and pulmonary gas exchange responses to a three minute bout of constant work rate exercise at 60 W, 90 W and 120 W were studied before then 16 and 23 months after posterior fossa decompression. The VE response to the three different levels of exercise was dramatically blunted so that the expected vertical relationship between PET(CO(2)) and VE during moderate exercise was replaced by an almost horizontal relationship with a slope ranging from 0.15 to 0.17 l/min/Torr. The reduced VE response was associated with a total lack of respiratory sensation during and following the exercise bouts. This abnormal ventilatory response to exercise persisted despite posterior fossa decompression. There was however no evidence of an alteration of the corticospinal pathway. Indeed, not only was there no sign of motor deficit but the patient was able both to mobilize 96% of her expected vital capacity and to voluntarily increase her ventilation to the level expected in a normal subject during exercise. This observation suggests that during exercise, motor control of respiratory muscles via a direct corticospinal pathway does not play a major role in adjusting phrenic motoneuron activity to the magnitude of the motor inputs to the exercising skeletal muscles.