BACKGROUND The objective evaluation of the hearing threshold below 1kHz by means of early auditory potentials leads to problems caused by the discrepancy between the required steep rise edge of stimulus and the frequency specificity of the potentials. Therefore the aim of this study was to evaluate the potential role of the AMFR as a diagnostic tool for the assessment of hearing below 1 kHz. METHODS The threshold of the AMFR was compared to the behavioral threshold in 13 normal hearing subjects and 46 patients with hearing loss. The stimulus used was an amplitude-modulated tone at the carrier frequencies of 0.5 and 1 kHz, a modulation frequency of 40 Hz; the modulation depth was 80%. The introduction of an empiric detectibility criterion based on spectral analysis of the response curve allowed the investigator to minimize the examination time and objectify the interpretation of the response. Additional investigations by means of highpass-masking took place to estimate the frequency specificity of the AMFR. RESULTS The results show a good correspondence of the AMFR threshold to the behavioral threshold. Depending on hearing loss the mean values of differences between AMFR threshold and behavioral threshold are 3 dB - 13 dB. The degree of correspondence was highest in the patients with the most severe hearing loss. An influence of underlying cause of the hearing loss could not be found. Additionally the frequency specificity of the response potential was proven with high-pass masking in normal hearing subjects. Masking with cut-off frequencies above the carrier frequencies had no influence on the response while masking at the carrier frequency resulted in a strong reduction of the response curve. CONCLUSIONS The results show that the 40 Hz-AMFR is a suitable method for the objective frequency-specific assessment of hearing in adults. Problems in the clinical use of the AMFR are caused by the long investigation time and the dependence of the potentials on the state of wakeness.