Hypercapnia attenuates the effects of static airway pressure (Paw) on phrenic burst frequency (f) and the expiratory duration (TE) in chloralose-urethan-anesthetized dogs. Surgical removal of the carotid bodies abolishes this interaction. Since halothane anesthesia in hyperoxia greatly impairs peripheral chemoreflexes, experiments were conducted to determine whether hypercapnia would attenuate the effects of Paw on f and TE in halothane-anesthetized dogs (approximately 1.5 minimum alveolar concentration). Integrated activity of the phrenic nerve was monitored as a function of Paw (2-12 cmH2O) in a vascularly isolated left lung at varied levels of arterial PCO2 (PaCO2; 38-80 Torr) controlled by inspired gas concentrations ventilating the denervated but perfused right lung. Halothane was administered only to the right lung. The results were as follows: 1) integrated phrenic amplitude increased with PaCO2 but was unaffected by Paw; 2) f decreased as Paw increased but was not affected by PaCO2; 3) the inspiratory duration (TI) increased as PaCO2 increased but was unaffected by Paw; 4) TE increased as Paw increased but was unaffected by PaCO2; and 5) there was no phrenic response to intravenous sodium cyanide (50-100 micrograms/kg). Thus, unlike chloralose-urethan-anesthetized dogs, hypercapnia does not attenuate the effect of lung inflation on f or TE in halothane-anesthetized dogs. Furthermore, hypercapnia increases TI during halothane anesthesia, an effect found after carotid denervation but not found in intact chloralose-urethan-anesthetized dogs. It is suggested that these differences between chloralose-urethan- and halothane-anesthetized dogs may be due to functional carotid chemoreceptor denervation by halothane.