This study was performed for the purpose of determining whether or not evoked otoacoustic emissions are useful as a clinical test. Two hundred and twenty-six sequences of the emission in response to stimulus tone bursts were averaged. The detection threshold of the emission was elevated in ears of inner ear impairment with profound sensorineural hearing loss, such as inner ear anomaly, mumps deafness, or sudden deafness, but it was not observed in ears of functional deafness. The mean interaural differences of emission threshold were near 35 dB in unilateral inner ear impairments with profound hearing loss. There was a positive correlation between the interaural difference of audiometric threshold and that of emission threshold in sudden deafness ears with various degrees of hearing loss. The incidence of continuous emission, whose duration was longer than 6 msec, was 30% in normal hearing ears and it was close to 90% in ears with bilateral or unilateral dip type hearing loss. The result was verified in a survey of a junior high school brass band. The conclusion is that there is clinical usefulness for the evoked otoacoustic emissions in evaluating cochlear function and in predicting noise susceptibility.