Anesthesia is known to affect the auditory brainstem response (ABR) in animals often used in hearing research. This study describes the differences in ABRs between awake and anesthetized FVB/N mice. Intracranial electrodes connected to a head fixation pedestal were used for click-evoked ABR recordings. This pedestal served to immobilize mice, either awake or under anesthesia, in a 'free' sound field. The presence of myogenic noise in the awake condition obviously increases recording time. However it is demonstrated that recording times can be significantly reduced by increasing the stimulus repetition rate from 23 up to 80 impulses per second. This causes only a small but significant increase in absolute peak latencies in the awake condition, but has no significant effect on the overall ABR-waveform, nor on the ABR-threshold, nor on the ABR interpeak latencies, nor on the absolute peak latencies in the anesthetized condition. Anesthesia with ketamine/xylazine caused a significant prolongation of ABR-peak latencies and interpeak latencies as well as a significant upward shift (8.0+/-1.8 dB) of ABR-thresholds as compared to the awake condition. Under anesthesia the measurement accuracy of peak latencies, interpeak latencies and thresholds decreases. In conclusion, the awake condition is preferable for more accurate measurements of ABR characteristics, in spite of the myogenic noise concomitant with this condition.