PROJECT SUMMARY Sleep is critically important for multiple forms of memory. Disrupted sleep, a hallmark of numerous neuropsychiatric disorders, is associated with debilitating cognitive symptoms, including problems with learning and memory. Furthermore, specific deficits in sleep-dependent procedural learning have been identified in patients with depression and schizophrenia. In depressed patients this effect has been doubly dissociated from sleep-dependent declarative learning, suggesting that distinct processes govern the role of sleep in declarative vs procedural learning. The overall goal of this proposal is to identify mechanisms whereby sleep contributes to the learning and maintenance of procedural memories. Although hippocampal sleep replay has been shown to exert a functional effect on learning, it is unknown how replay alters neural encoding in order to affect behavior. Furthermore, it remains unclear how sleep exerts its effects on non-hippocampal memory, which is thought to include most types of motor, procedural, and perceptual memories. The songbird system offers a valuable opportunity to study a clear form of sequential motor sleep replay in the context of a learned behavior (song) that is both natural and highly stereotyped. Motor regions in the songbird forebrain exhibit replay of song- associated activity patterns during sleep. Moreover, sleep leads to changes in motor encoding, and exerts behavioral effects on song learning. We propose to study links between sleep, replay, and behavior in both adult birds maintaining their song and juvenile birds learning to sing. In adult birds, we will use electrophysiology, polysomnography, electromyography, and sophisticated computational methods to identify the timing and content of normal replay events and their relationship with sleep architecture. Harnessing a well- known property of the song system, we will evoke replay-like events to track changes in motor coding across the night. We will then perturb auditory feedback and examine the response of sleep replay and sleep architecture. In juvenile birds, we will experimentally control song learning and explore the role of sensory feedback in a known sleep-dependent behavioral effect on song learning. Finally, we will use a combination of pharmacology and electrophysiology to identify circuit mechanisms of the appearance of proto-replay, which has been linked to the onset of sleep-dependent learning effects. The results of this work will elucidate circuit mechanisms linking sleep replay with behavior and will clarify the role of sensory feedback in sleep-dependent procedural learning. These insights will help pave the way for developing interventions into the many neuropsychiatric illnesses in which sleep and learning are disrupted, and for harnessing sleep processes to ameliorate deficits in learning.