Tracheal gas insufflation (TGI) improves the efficiency of CO2 elimination by replacing CO2 in the anatomic dead space proximal to the catheter tip with fresh gas during expiration. Turbulence generated by gas exiting the catheter tip may also contribute to alveolar ventilation. To separate distal (turbulence-related) and proximal (washout of dead space) effects of TGI, we compared the efficacy of a straight and an inverted catheter during continuous and expiratory TGI in six mechanically ventilated dogs. We reasoned that the inverted catheter cannot improve CO2 elimination from more distal conducting airways. During continuous TGI with the straight catheter, arterial PCO2 (PaCO2) decreased significantly from baseline (without TGI) of 56 +/- 10 Torr to 38 +/- 8, 36 +/- 8, and 35 +/- 8 Torr at catheter flow rates (Vcath) of 5, 10, and 15 l/min, respectively. For the same conditions, PaCO2 was always higher (P < 0.001) with the inverted catheter (42 +/- 10, 41 +/- 10, and 41 +/- 10 Torr). PaCO2 was lower with the straight (40 +/- 9 Torr) than with the inverted catheter (44 +/- 10 Torr, P < 0.001) during TGI delivered only during expiration at a Vcath of 10 l/min. End-expiratory lung volume relative to baseline increased during continuous, but not during expiratory, TGI and was significantly greater with the straight than with the inverted catheter (P < 0.0001). Our data confirm that the primary mechanism of TGI is expiratory washout of the proximal anatomic dead space but also suggest a minor contribution of turbulence beyond the tip of the straight catheter.