The sensitivities of current monitoring methods for detection of air embolism were compared in eight anesthetized dogs. Air was infused at controlled rates of 0.001 and 0.005 ml. k−1.min−1for 1 min; 0.01, 0.05, 0.1, 0.2, and 0.4 ml.kg−1for 6 min; and 5 ml. kg−1bolus injection. Based on the mean quantity of air infused to elicit a positive response, the monitors could be placed into three significantly different sensitivity groups. Transesophageal echocardiography (TEE) and precordial Doppler ultrasound were the most sensitive monitoring methods detecting 0.19 and 0.24 ml. kg−1of air, respectively. TEE detected air during six infusions in which the Doppler failed to do so. The next most sensitive group of monitoring methods included pulmonary artery pressure (PAP), end-tidal CO2(Ptco2), arterial oxygen tension (Pao2), and transcutaneous oxygen tension (Ptco2). The mean quantity of air infused to elicit a positive response in this group of monitors ranged from 0.61 to 0.76 ml. kg−1. The response of Ptco2, Pao2, PETco2, and PAP equally reflected the quantity of air infused. The least-sensitive group of methods included arterial and transcutaneous carbon dioxide tension and systemic arterial blood pressure. These data indicate that TEE is more sensitive than Doppler ultrasound and that PAP, PETco2, and Ptco2are equally sensitive in detecting venous air embolism in the dog.