The phenomenon of conditioning-related protection, whereby prior exposure to moderate-level, non-traumatic, sound protects the ear from subsequent traumatic exposure, has been documented in a number of mammalian species. To probe the molecular mechanisms underlying this effect, the mouse would be a useful model; however, a previous study reported no conditioning effects in this species (Fowler et al. , 1995). In our study, mice (CBA/CaJ) were exposed to a traumatic octave-band noise (8-16 kHz at 100 dB SPL for 2 h) with, or without, prior exposure to a sound-conditioning protocol consisting of exposure to the same noise band at lower sound pressure levels. Two conditioning protocols were investigated: one (81 dB SPL for 1 week) was analogous to those used in other conditioning studies in mammals; the second was significantly shorter (89 dB SPL for 15 min). Noise-induced permanent threshold shift (PTS) was assessed in a terminal experiment, after the traumatic exposure, via compound action potentials. Neither conditioning protocol elevated threshold, indeed both protocols increased amplitudes of distortion product otoacoustic emissions when animals were conditioned but not traumatized. Both conditioning exposures significantly reduced PTS from the subsequent traumatic exposure, compared to groups exposed without prior conditioning. Protective effects of 15-min conditioning were maximal when the condition-trauma interval was 24 h; protection disappeared when the traumatic exposure was presented 48 h after conditioning. These data are consistent with the view that protein synthesis is required for expression of the protective effect. The enhancement of distortion products in the condition-only state suggests that conditioning changes outer hair cell function.