Anxiety can induce hyperventilation, and the resultant hypocapnia and hypocalcaemia can lead to paraesthesiae and tetany. To define the nature of the disturbance created in peripheral nerve, the excitability of cutaneous and motor axons was monitored in 6 normal subjects requested to hyperventilate until paraesthesiae developed in the hands, face and trunk. This occurred when alveolar PCO2 (PACO2) had declined on average by 20 mmHg. Spontaneous EMG activity developed when PACO2 had declined by a further 4 mmHg. Changes in the excitability of cutaneous and motor axons were measured from changes in the compound action potentials evoked by stimulation of the median nerve at the wrist and recorded over the digital nerves of the index finger and over the thenar muscles, respectively. As PACO2 declined, the size of the compound sensory and muscle potentials evoked by a constant stimulus progressively increased, indicating an increase in axonal excitability. These changes occurred before paraesthesiae or tetany developed. In each subject there was a statistically significant inverse correlation between PACO2 and axonal excitability. Independent of this increase in axonal excitability, there was no significant change in the supernormal phase of the recovery cycle of cutaneous axons. Microneurographic recordings from the median nerve in 2 subjects revealed spontaneous bursting activity of cutaneous axons, perceived as paraesthesiae. It is concluded that the paraesthesiae and tetany induced by hyperventilation result solely from changes in the excitability of cutaneous and motor axons in the peripheral nerve, presumably due to an alteration in the electrical properties of the axonal membrane resulting from a reduced plasma [Ca2+]. The supernormal phase may entrain the ectopic discharge and thereby determine the maximal discharge frequency of impulses in ectopically generated trains, but does not otherwise contribute to the physiological disturbance.