Thirty normal-hearing college-age adults were tested by standard and respiration audiometry (RA) at 1000 Hz to investigate the effects of subject knowledge of the RA procedure. Three levels of sophistication were employed: (1) no knowledge, (2) full knowledge, and (3) deception. No significant effects on RA thresholds were revealed as a function of subject knowledge. RA itself was found to be less satisfactory as a means of measuring auditory threshold because of the difficulty in obtaining responses from some subjects and the degree of subjectivity required to judge responses.