Tracheal gas insufflation (TGI) has been shown to be a useful adjunct to mechanical ventilation, decreasing PaCO2 during permissive hypercapnia. While TGI can be used either with pressure (PCV) or volume-controlled ventilation and continuously or only during the expiratory phase (Ex-TGI), there are no controlled studies evaluating the effects of Ex-TGI with PCV in acute lung injury when the direction of the insufflated flow or the inspiratory:expiratory (I:E) ratio are varied. We evaluated the effect that Ex-TGI with PCV would have on CO2 removal during both direct and reverse insufflated flow direction with varied I:E ratios when peak airway pressure, total positive end-expiratory pressure (PEEP), and tidal volume (VT) were kept constant. In addition we examined the effect that insufflation flow directed toward the mouth (reverse flow) would have on the generation of PEEP compared with flow directed toward the carina (direct flow). After saline lavage, nine sheep were ventilated with PCV to a baseline PaCO2 of 80 mm Hg. Ex-TGI (10 L/min) was then randomly applied in the reverse and direct direction with I:E set at 1:2 or 2:1. During 1:2 I:E PaCO2 decreased from 78 +/- 4 mm Hg to 60 +/- 7 mm Hg (23.5 +/- 8.9%) with direct flow and to 64 +/- 5 mm Hg (18.5 +/- 5.5%) with reverse flow (p < 0.05), whereas during 2:1 I:E PaCO2 decreased from 80 +/- 4 mm Hg to 69 +/- 8 mm Hg (13.7 +/- 9.2%) with direct flow and to 66 +/- 4 mm Hg (17.2 +/- 4.4%) with reverse flow (p < 0.05). Greater PEEP was developed with direct flow (2.8 cm H2O I:E 1:2 and 4.0 cm H2O I:E 2:1) than with reverse flow (-0.9 cm H2O I:E 1:2 and -0.4 cm H2O I:E 2:1), p < 0.05. There was no difference in the PaCO2 change between I:E with reverse flow, but the PaCO2 decrease was greater (p < 0.05) during 1:2 versus 2:1 I:E with direct flow. CO2 removal during PCV and Ex-TGI is more consistent with reverse flow than with direct flow and PEEP level is less affected by TGI with reverse flow than with direct flow.