Auditory brain-stem responses (ABRs) were measured in the awake state and with ketamine and xylazine anesthesia in adult gerbils. Surface recorded vertex-positive components of the ABR were analyzed with respect to the awake and anesthetized states as a function of stimulus frequency. ABR thresholds were not altered with ketamine/xylazine. Small increases in peak latency were associated with anesthesia for all components except wave P1. Increases in absolute latency were progressively greater for successive peaks, reaching an average shift of 0.41 msec for wave P6. Amplitude changes with anesthesia were more variable, with increases generally seen for waves P4 and P6. Significant anesthesia effects on peak latency and amplitude were independent of stimulus frequency. These data confirm previously reported ABR sensitivity to non-barbiturate anesthesia. Direct comparisons of ABR interpeak intervals or amplitude ratios from awake versus anesthetized animals must account for the effects of barbiturate and non-barbiturate agents. However, the stability of response threshold and the small magnitude of latency and amplitude changes with a ketamine and xylazine regimen demonstrate that accurate electrophysiological measures of hearing sensitivity and auditory brain-stem activity can be obtained in anesthetized animals, provided that temperature and other parameters are maintained within normal physiological limits.