The ascending auditory pathway consists of the auditory periphery, including mechanosensitive hair cells in the cochlea, spiral ganglion neurons forming the auditory nerve, and at least five stations until sound reaches the auditory cortex. All synapses along this pathway require proper expression and function of presynaptic Ca2+ channels for triggering transmitter release. Voltage-gated calcium channels (VGCCs) are protein complexes composed of an α1 pore-forming subunit and auxiliary subunits β and α2δ. VGCCs of cochlear inner hair cells (IHCs) are mainly composed of the α1 subunit Cav1.3 (contributing to ~90% of ICa; refs. 1,2) and β2 (contributing to ~70% of ICa; ref. 3), and lack of Cav1.3 or β2 causes deafness. Auxiliary subunits assist in trafficking and membrane surface expression of VGCCs and modulate their biophysical properties. Recently, novel roles for α2δ in synapse development have been suggested (refs. 4,5). So far, expression and contribution of the four α2δ subunits α2δ1-4, and their specificity for the pore-forming α1 and auxiliary β subunits are still unknown. Because quantitative PCR revealed the expression of α2δ2 in IHCs at pre-hearing and hearing age, we here analyzed hearing and IHCs of ducky mice (α2δdu/du), a mouse line with a mutation in the cacna2d2 gene encoding a non-functional α2δ protein. Auditory brainstem response (ABR) click thresholds were increased by 18.3 dB in the mutants, and frequency ABR showed a significant elevation of hearing thresholds by about 20 dB between 11.3 and 22.6 kHz (n ≥ 11 ears). Distortion-product otoacoustic emission (DPOAE) amplitudes were slightly larger in α2δdu/du compared with wild-type animals, indicating normal outer hair cell (OHC) function. Synaptic morphology, as assessed by immunolabelling, revealed normal expression of presynaptic Cav1.3 and β2 as well as postsynaptic GluR2/3-proteins in close apposition with synaptic ribbons in IHCs of α2δdu/du mice. However, both peak Ca2+ and peak Ba2+ current densities of α2δdu/du IHCs were reduced by 29 % (n ≥11). Moreover, the voltages for half-maximum activation of Ca2+ and Ba2+ currents were shifted to more positive potentials by 5 mV and 7 mV, respectively, whereas the voltage sensitivities (steepness of the activation curves) were unaffected. Inactivation of Ca2+ and Ba2+ currents over 300 ms depolarization steps was unchanged. These results are in accordance with the classical role of α2δ subunits in surface expression and gating modulation of VGCCs. In wild-type IHCs, Cav1.3 and β2 appear to co-assemble with α2δ2. Lack of α2δ2 cannot be fully compensated, because of reduced IHC Ca2+ currents causing a moderate hearing impairment. We detected another form of hearing impairment in a mouse line carrying a targeted deletion of α2δ3 (α2δ3-/-; ref. 6). ABR hearing thresholds were slightly but significantly higher in α2δ3-/- mice by 5 dB and 8-19 dB for click and pure tone stimuli, respectively, compared with wild-type littermates. Additionally, ABR waveforms showed a reduction in the amplitude of wave II and distortion of waves III and IV in α2δ3-/-, suggesting impaired signal transmission along the auditory pathway. In behavioral experiments, α2δ3-/- mice learned to discriminate pure tones but completely failed to discriminate temporally structured amplitude-modulated tones, as demonstrated by an aversive go/no-go shuttle-box paradigm. IHCs of α2δ3-/- mice had robust Ba2+ currents through Ca2+ channels and normal numbers of Cav1.3-immunopositive spots, which were co-localized with synaptic ribbons. Together with normal DPOAE amplitudes in α2δ3-/- mice, these data suggest that deletion of α2δ3 does not impair IHC and OHC function. α2δ3 expression was predominant in spiral ganglion neurons and in auditory brainstem nuclei. Impaired synaptic transmission from auditory nerve fibers to bushy cells was confirmed by juxtacellular in-vivo recordings of sound-evoked activity, which revealed shallower rate-level functions and increased latencies at these synapses in α2δ3-/- mice. Sizes of auditory nerve fiber synaptic boutons, which terminate at the somata of bushy cells, were significantly smaller in mutants, and the level of Cav2.1 channel immunoreactivity was reduced. These results identify a novel role for the α2δ3 auxiliary subunit in establishing and maintaining the morphology and function of specific synapses in the mammalian auditory pathway and identify the α2δ3 gene as a potential candidate gene for an auditory processing disorder.