Ten sulphur-crested cockatoos (Cacatua galerita) were anesthetized twice, 2 weeks apart, to evaluate the effects on respiratory function of endotracheal (ET), caudal thoracic (CT) air sac, and clavicular air sac administration of isoflurane anesthesia. Four administration methods were evaluated by administering anesthetic gases through either the ET tube or the air sac as an open or closed system. Respiratory function was monitored by arterial and venous blood gas determination, gas-in-gas determination of both ET tube and air sac gases, and spirometry. Both ET tube and CT air sac administration provided a reliable method of maintaining anesthesia and resulted in minimal alteration in respiratory function. Blood pH, partial pressure of carbon dioxide (Pco₂), bicarbonate, total carbon dioxide, and base excess were not significantly altered by CT air sac administration. Increased partial pressure of oxygen and oxygen saturation were evidence of hyperoxia. CT air sac Pco₂ decreased to 12.7 ± 7.9 mm Hg with CT administration, and this was sufficient to maintain spontaneous respiratory movements. Spirometry demonstrated maintenance of tidal and minute volumes with each administration method despite apparent reduction in visible respiratory movements. Clavicular air sac administration was not successful in providing ventilation or maintaining anesthesia. Despite the demonstrated ability to flush a bolus of gas through this air sac and out the ET tube, no spontaneous ventilation was detected in this air sac in either an open or closed system, preventing delivery of sufficient isoflurane to maintain anesthesia.