The effects of a 90-min infusion of somatostatin (1 mg/h) on ventilation and the ventilatory responses to hypoxia and hypercapnia were studied in six normal adult males. Minute ventilation (VE) was measured with inductance plethysmography, arterial 02 saturation (SaO2) was measured with ear oximetry, and arterial PCO2 (Paco2) was estimated with a transcutaneous CO2 electrode. The steady-state ventilatory response to hypoxia (delta VE/delta SaO2) was measured in subjects breathing 10.5% O2 in an open circuit while isocapnia was maintained by the addition of CO2. The hypercapnic response (delta VE/delta PaCO2) was measured in subjects breathing first 5% and then 7.5% CO2 (in 52-55% O2). Somatostatin greatly attenuated the hypoxic response (control mean -790 ml x min-1.%SaO2 -1, somatostatin mean -120 ml x min-1.%SaO2 -1; P less than 0.01), caused a small fall in resting ventilation (mean % fall - 11%), but did not affect the hypercapnic response. In three of the subjects progressive ventilatory responses (using rebreathing techniques, dry gas meter, and end-tidal Pco2 analysis) and overall metabolism were measured. Somatostatin caused similar changes (mean fall in hypoxic response -73%; no change in hypercapnic response) and did not alter overall O2 consumption nor CO2 production. These results show an hitherto-unsuspected inhibitory potential of this neuropeptide on the control of breathing; the sparing of the hypercapnic response is suggestive of an action on the carotid body but does not exclude a central effect.