To assess the incidence of posthyperventilation apnea (PHA), breathing patterns after active voluntary hyperventilation were determined during EEG recording by means of: (1) a thermocouple in 1060 patients; (2) a pneumotachograph coupled with a capnograph in 100 further patients. All the patients were randomly chosen. PHA of 12.6 +/- 0.48 occurred in only 18% of awake subjects and was not related to the magnitude of the lowering of end-tidal FECO2 during the test. PHA frequency increased when sleep followed hyperventilation and in patients with clinical or EEG evidence of cerebral pathology and seemed to be age-related. This study supports the hypothesis that in man a central neural mechanism closely linked to vigilance and connected with active breathing supplies sufficient neural facilitation to prevent the apnea consequent on the decrease in chemical humoral stimulation. Posthyperventilation sleep reduces this neural facilitation, allowing the lack of chemical stimulation to induce apnea. The high incidence of PHA reported in some studies might be due to a decrease in vigilance which was not detected by the usual behavioral or electroencephalographic tests.