BACKGROUND: If intravenous access cannot be attained during resuscitation of adult patients, endotracheal application of at least 2 mg of adrenaline is recommended. However, the effects of this intervention have not yet been demonstrated in adults. OBJECTIVE: To demonstrate the effects of adrenaline administered through the airways. DESIGN: Prospective clinical trial. SETTING: Operating theater at university hospital. PATIENTS: 34 patients receiving implantable cardioverter defibrillators under general anesthesia. INTERVENTION: When mean arterial pressure decreased below 80 mm Hg, 100 times the effective central intravenous dose of adrenaline (mean +/- SD, 1.3+/-0.6 mg [range, 0.7 to 3 mg]) was administered over 5 seconds into the endotracheal tube or through a bronchial catheter. Ten forced ventilations followed. MEASUREMENTS: Hemodynamic variables were recorded with a polygraph recorder. Adrenaline levels were measured in 13 patients. RESULTS: Plasma levels and arterial pressure increased in all patients (P < 0.002). Higher plasma levels (P < 0.039) and greater arterial pressure (P < 0.001) were achieved with this method than with intravenous injection. The effects of adrenaline did not differ between the two airway routes. Sustained ventricular arrhythmia did not occur. CONCLUSION: These substantial effects support the standard recommendation to consider the airways as an alternate route for at least 2 mg of adrenaline during resuscitation.