Introduction: We evaluated the accuracy of transcutaneous CO 2 (tPCO 2 ) and end-tidal CO 2 (ETCO 2 ) monitors in a porcine model of Acute Lung Injury (ALI) due to smoke inhalation and burns. Methods: Anesthetized pigs (n=6) received smoke inhalation injury and 40% total body surface area (TBSA) full-thickness flame burns. They were then transferred to an animal ICU. A SenTec system was used for tPCO 2 measurements, and an Oridion Capnostream for ETCO 2 . PaCO 2 was analyzed with an i-STAT blood analyzer. These measurements were recorded every 6 hrs and as needed. Data are means ±SD, and were analyzed (Bland-Altman and linear regression) both pre- and post-ALI, as well as during hemodynamically stable and unstable (unstable = mean arterial pressure (MAP) < 60 mmHg or receiving i.v.norepinephrine). Results: See Table. All animals developed ALI (PaO 2 /FiO 2 <300) at 23.5±15.1 hrs. Mean survival time was 46.3 ± 26.3 hrs. Conclusions: In a porcine model of ALI and under hemodynamically stable conditions, tPCO 2 was an acceptable surrogate for PaCO 2 and was more accurate than ETCO 2 . Overall, tPCO 2 monitoring is a consistent noninvasive substitute for PaCO 2 . Its shortcomings include a slower response time to PaCO 2 changes than ETCO 2 .