Precordial ultrasonic Doppler devices are effective monitors for detecting venous air emboli (VAE). However, placing an ultrasonic probe on the anterior part of the chest of a prone patient can lead to dislodgment or pressure sores and makes the probe inaccessible to the anesthesiologist. The purpose of this study was to compare placement of a Doppler probe on the patient's back with the traditional precordial site for the ability to detect VAE. We enrolled infants and children undergoing neurosurgical procedures in the prone position in the study. After establishment of general anesthesia and endotracheal intubation, we applied an ultrasonic Doppler probe to the right sternal border of the patient's chest. Anterior insonation was performed with the patient in the supine position. Saline was rapidly injected to verify the efficacy of the monitor (injection test). The patient was turned to the prone position and we placed the Doppler probe between the right scapula and spine. Posterior insonation with saline injection was performed with the patient in the prone position. We obtained positive tests in all patients from the anterior site. Positive tests were obtained from the posterior site in 23 of 24 (96%) children < 10 kilograms (group I), 28 of 39 (72%) children between 10 and 20 kg (group II), and 6 of 22 (27%) children > 20 kilograms (group III). This study demonstrates that a posterior Doppler probe can be effective for monitoring infants at risk of VAE. However, this method is not reliable in children weighting > 10 kg.