OBJECTIVE To determine the ability of a disposable colorimetric CO2 detector to accurately confirm or refute endotracheal tube placement. METHODS Two hundred fifty prospective emergency intubations. METHODS Emergency intubations performed in the emergency department, helicopter, and prehospital ground environment. METHODS Intubations were performed by emergency medicine residents, paramedics, and flight nurses. METHODS The FEF CO2 detector was applied after 250 emergency intubations. Notation of color change indicating intratracheal placement was recorded in each case. Confirmation of refutation of the detector's results was determined subsequently through traditional methods. RESULTS The sensitivity for confirmation of endotracheal intubation in the 137 patients with a palpable pulse was 100%. However, only 76 of 103 patients (sensitivity, 72%) in cardiac arrest had endotracheal intubation confirmed by color change. The device was uniformly specific for tracheal intubation in 73 arrested patients in whom a color change was noted (100%). There was one instance (of a total of seven misintubations) in which a positive color change was noted, but the tube was not intratracheal (specificity, 86%). Overall sensitivity for tracheal intubation was 88% (95% confidence limits; range, 0.83 to 0.92), and specificity for tracheal intubation was 92% (95% confidence limits; range, 0.62 to 0.99). CONCLUSIONS The FEF colorimetric detector reliably detects intratracheal placement in the nonarrested patient. Its use in prolonged cardiac arrest merits further study.