Normoxic and hypoxic respiration has been measured in awake rats after denervation procedures designed to eliminate the regulatory input from the carotid bodies, from all chemosensory tissue supplied by the glossopharyngeal nerve (n. IX), and from abdominal chemoreceptors. Studies were made 1 day after section of the carotid sinus nerve (c.s.n.), n. IX (at a level including c.s.n.), the abdominal vagus (n. Xa) and combinations of these nerves. Results were compared with those found in normal controls. C.s.n. section led to hypoventilation in both normoxia and hypoxia, reductions in respiratory frequency being consistent and substantial, and reductions in tidal volume varying with the degree of hypoxia. By comparison, section of n. IX produced significantly greater reductions of both normoxic and hypoxic ventilation. Section of n. Xa produced no significant change in normoxic ventilation but in hypoxia produced a significant small reduction in ventilation, mostly from an effect on tidal volume. Denervation of all the associated chemosensory tissue by combined section of n. IX and n. Xa demonstrated a summation of effects but left two distinct residual responses, one to mild hypoxia, and one to severe hypoxia, both associated mainly with increases of tidal volume. The experiments demonstrate that glomus tissues at different sites in the rat produce significant and distinct contributions to respiratory regulation. Denervation of all known receptors shows that significant ventilatory responses to hypoxia are still produced, either by unrevealed peripheral chemoreceptors, or by central neural mechanisms.