Neural activity has been implicated as having both a trophic function and a role in synaptic specificity. Sensory deprivation studies in a large number of developing systems have resulted in the pathological morphology of neurons and abnormal response properties. If the relative timing of discharge among afferent terminals is a cue employed by the developing system to refine the array of synaptic connections, then altering the discharge patterns should hinder this process. In the present experiments, we investigate the role played by the temporal pattern of neural activity during the ontogeny of frequency tuning in the mouse central auditory system. Postnatal animals were exposed to acoustic stimuli, repetitive clicks, that continuously entrained a large proportion of primary afferents from the onset of hearing until an age at which tuning curves should have been adult-like. The amount of fatigue to repetitive clicks was characterized at the level of the eighth nerve and inferior colliculus in normal animals. Frequency tuning curves obtained from the inferior colliculus were used as an assay for the specificity of neural connections. Click-reared animals had significantly broader tuning curves than did normally reared mice, particularly for units with best frequencies in the 10- to 15-kHz range. Furthermore, it was found that this change could not be attributed to the selective loss of the sustained component of the response. The affected range is interpreted in terms of the frequency spectrum of the click and the fact that lower frequency regions of the inferior colliculus were found to habituate rapidly. The click-rearing environment did not appear to affect unit spontaneous activity or response latency, nor did it alter the tonotopic map in the inferior colliculus. We argue against the possibility of cochlear damage based on threshold and high frequency cutoff measurements. Mice were reared in a second acoustic environment, repetitive pulses of two added frequencies, as a control for the effects of the click stimulus. This rearing paradigm did not lead to a broadening of tuning curves. It did, however, alter the properties of bimodal tuning curves. For units with bimodal tuning curves having best frequencies in the range of the rearing frequencies, it was found that the second excitatory area had a lower than normal threshold. In addition, the frequency range separating the peaks of the two excitatory regions was statistically smaller. These results are discussed with reference to the specific frequencies used in the rearing paradigm.