The fast brain stem response can be recorded even at the threshold; it informs on hearing in the high frequency range of greater than or equal to 2000 cps. Especially in those frequencies often the differentiation between sensorial and neural hearing impairment is of interest--f. e. controlling the pure tone threshold in noise: at sensorial hearing loss tones remain heard within noise of equal loudness unaltered, at neural hearing loss the tone vanishes in noise, e. g. the tone-intensity must be increased to make the tone arise again. This phenomenon known from subjective audiometry mentally can be transferred to the recording of acoustic evoked potentials and can be proved by the behaviour of latency of P6: at sensorial hearing loss the latency is prolonged (corresponding to minus 20 dB) but the response furthermore is recorded in the noise as well as at the click masked threshold; at neural hearing loss, however, the response can be recorded again with a click-intensity, which exceeds that of the noise clearly.