Research in animal models has demonstrated that electrical stimulation from a cochlear implant (CI) may help prevent degeneration of the cochlear spiral ganglion (SG) neurons after deafness. In cats deafened early in life, effective stimulation of the auditory nerve with complex signals for several months preserves a greater density of SG neurons in the stimulated cochleae as compared to the contralateral deafened ear. However, SG survival is still far from normal even with early intervention with an implant. Thus, pharmacologic agents and neurotrophic factors that might be used in combination with an implant are of great interest. Exogenous administration of GM1 ganglioside significantly reduces SG degeneration in deafened animals studied at 7-8 weeks of age, but after several months of stimulation, GM1-treated animals show only modestly better preservation of SG density compared to age-matched non-treated animals. A significant factor influencing neurotrophic effects in animal models is insertion trauma, which results in significant regional SG degeneration. Thus, an important goal is to further improve human CI electrode designs and insertion techniques to minimize trauma. Another important issue for studies of neurotrophic effects in the developing auditory system is the potential role of critical periods. Studies examining animals deafened at 30 days of age (rather than at birth) have explored whether a brief initial period of normal auditory experience affects the vulnerability of the SG or cochlear nucleus (CN) to auditory deprivation. Interestingly, SG survival in animals deafened at 30-days was not significantly different from age-matched neonatally deafened animals, but significant differences were observed in the central auditory system. CN volume was significantly closer to normal in the animals deafened at 30 days as compared to neonatally deafened animals. However, no difference was observed between the stimulated and contralateral CN volumes in either deafened group. Measurements of AVCN spherical cell somata showed that after later onset of deafness in the 30-day deafened group, mean cell size was significantly closer to normal than in the neonatally deafened group. Further, electrical stimulation elicited a significant increase in spherical cell size in the CN ipsilateral to the implant as compared to the contralateral CN in both deafened groups. Neuronal tracer studies have examined the primary afferent projections from the SG to the CN in neonatally deafened cats. CN projections exhibit a clear cochleotopic organization despite severe auditory deprivation from birth. However, when normalized for the smaller CN size after deafness, projections were 30-50% broader than normal. After unilateral electrical stimulation there was no difference between projections from the stimulated and non-stimulated ears. These findings suggest that early normal auditory experience may be essential for the normal development (or subsequent maintenance) of the topographic precision of SG-to-CN projections. After early deafness, the CN volume is markedly smaller than normal, and the spatial precision of SG projections that underlie frequency resolution in the central auditory system is reduced. Electrical stimulation over several months did not reduce or exacerbate these degenerative changes. If similar principles pertain in the human auditory system, then findings in animal models suggest that the basic cochleotopic organization of neural projections in the central auditory system is probably intact even in congenitally deaf individuals. However, the reduced spatial resolution of the primary afferent projections in our studies suggests that there may be inherent limitations for CI stimulation in congenitally deaf subjects. Spatial (spectral) selectivity of stimulation delivered on adjacent CI channels may be poorer due to the greater overlap of SG central axons representing nearby frequencies. Such CI users may be more dependent upon temporal features of electrical stimuli, and it may be advantageous to enhance the salience of such cues, for example, by removing some electrodes from the processor "map" to reduce channel interaction.