Project Summary / Abstract Binaural cues enable sound localization, which supports active listening by spatially segregating competing sound streams. Consequently monaural hearing loss impairs sound localization and listening comprehension, despite normal function in the spared ear. Encouragingly, humans and animals with monaural hearing loss can re-learn sound localization if provided with appropriate training, though success is variable. Identifying actionable targets to improve outcomes could thus benefit hearing health, but progress is limited by our meager understanding of central plasticity mechanisms that support learned sound localization. Interestingly, descending (“corticofugal”) projections from auditory cortex to subcortical auditory centers are required for monaurally deprived animals to relearn sound localization. However, their precise role in this process is poorly understood. This renewal addresses this knowledge gap by building upon our recent progress studying the role of corticofugal projections in discriminative learning. In tandem with unpublished data, our results suggest a new working hypothesis: Auditory corticofugal neurons and/or their subcortical targets are plasticity loci to establish learned, conjunctive representations of sounds and their behaviorally relevant consequences. Such “mixed selectivity” is powerful: It enables linking “where” and “what” information across time, thereby generating context-specific representations of auditory space that could support learned sound localization. To test this hypothesis, we have developed a new sound localization task that head-fixed mice learn with binaural cues, and can re-learn following monaural conductive hearing loss. We will combine this approach with longitudinal Ca2+ imaging and optogenetics to record and manipulate activity in corticofugal neurons and their subcortical targets. The results will provide new insights into central auditory plasticity during learned sound localization, w