This study was conducted to assess the CO2-elimination efficiency of tracheal gas insufflation (TGI) in 20 patients with acute respiratory distress syndrome and to compare its efficacy during volume-controlled (VCV) and pressure-controlled ventilation (PCV). TGI was initially applied as an adjunct to VCV, with continuous flows (Vcath) of 4 and 6 L/min delivered through a catheter positioned 2 cm above the carina. Total effective tidal volume (VTeff) was held constant. The percent reductions in PaCO2 (% delta PaCO2) were 13.3 +/- 2.1 and 16.7 +/- 2.7% at Vcath 4 and 6 L/min, respectively, which correlated well with the percent reduction in the end-tidal PCO2 from baseline (% delta PETCO2) (r = 0.75, p < 0.001). An inverse relationship (r = -0.65, p < 0.001 and r = -0.59, p < 0.01 at Vcath 4 and 6 L/min, respectively) was found between the % delta PaCO2 and the baseline ratio of artery to PETCO2 difference to PaCO2, which was determined as the fraction of alveolar dead space (VDalv) relative to total alveolar ventilation. Twelve patients were subsequently switched to PCV combined with Vcath 6 L/min, which provided a % delta PaCO2 of 16.1 +/- 3.0% (p = NS versus 17.1 +/- 2.6% during VCV). These data suggest that in patients with ARDS the change in PETCO2 may be helpful in predicting the decrement in PaCO2 during TGI, and the existence of a high VDalv tends to limit its effectiveness. Further, the efficacy of TGI with VCV is equivalent to that with PCV.