A clinical study was conducted to determine the extent to which sounds generated by air insufflations through feeding tubes could be used to predict where the tubes' ports ended in the gastrointestinal tract (esophagus, stomach or proximal small intestine), and to differentiate between gastric and respiratory placement. Sounds generated by a series of air insufflations through the tubes of 85 acutely-ill adult subjects were recorded. One hundred fifteen usable tape-recordings of sound sequences were obtained. The principal investigator later played the tapes for the research team members (who were also skilled clinicians) so that they could independently record their impressions of the sounds. (No identification of subjects or tube positions were given on the tapes; however, each site of auscultation was announced.) The results were collated and compared against the hypotheses and actual tube location (as determined by X-ray). Overall, the average percentage of correct classifications of each tape was 34.4%. Descriptive data were reported for three subjects with feeding tubes inadvertently positioned in the respiratory tract; air insufflations were clearly audible in 2 of the 3 cases. Clinicians should not rely on the auscultatory method to differentiate gastric from intestinal placement, nor gastric from respiratory placement of feeding tubes.