Abstract
Repeat within ear threshold variance can be considered to be made up of variables associated with the sensory system and those associated with the criterion adopted by the subject as to what is his threshold. Conventional audiometry is not able to separate out the effects of these different sources. Signal detection theory as applied to psychophysics would seem to offer a possible method of measuring the effects of the sensory and criterial variables separately. The variance in the measures of each of these terms for normal ears should be smaller than the variance of conventional threshold determinations. Hence the behaviour of normal ears can be more precisely specified and so departures from normality should be detected earlier than is at present possible. The problems which must be overcome if signal detection theory is to be successfully adapted for use in routine audiometry are discussed.