The effects of changes in static airway pressure (Paw) and arterial PCO2 (PaCO2) on phrenic nerve activity were studied in unanesthetized, decerebrate dogs and compared with previous results from chloralose/urethane anesthetized dogs using the same experimental preparation (Mitchell et al. 1982; Mitchell and Selby 1987). In ten mid-collicular decerebrate dogs, the lungs were independently ventilated while the left pulmonary artery was occluded and the right vagus nerve was transected. Changes in left lung Paw, therefore, exerted effects on pulmonary stretch receptors without altering blood gases; changes in the inspired gas ventilating the right lung controlled blood gas composition, without altering lung volume feedback. Phrenic burst frequency (f) and integrated amplitude (Phr) were monitored while Paw was varied between 2 and 12 cmH2O at various constant levels of PaCO2 between 31 and 69 mmHg. The major findings of this study are: (1) hypercapnia decreases the slope of the relationship between expiratory duration (tE) and Paw in both decerebrated and anesthetized dogs; (2) hypercapnia increases the inspiratory duration (tI) in decerebrated, but not anesthetized dogs; and (3) hypercapnia decreases the slope of the relationship between f and Paw due to these effects on tE and tI. These results support previous studies indicating that vagal and suprapontine mechanisms exert independent effects on respiratory timing. It is concluded that neither suprapontine influences nor anesthesia are necessary in the mechanism underlying interactions between stretch receptors and CO2-chemoreceptors in modulating tE. Furthermore, decerebration reveals a unique effect of CO2-chemoreceptors on tI, an effect found in anesthetized dogs only after carotid denervation.