Abstract Sleep dysregulation is a hallmark of alcohol use disorders (AUDs) and disrupted sleep can contribute to relapse even after months of abstinence. Despite the well-recognized sleep-AUD interactions, few studies have investigated how sleep changes over the development of excessive drinking, and the mechanisms by which sleep disruptions exacerbate drinking and/or relapse liability are largely unknown. Based on several key findings from published studies and preliminary work, one hypothesis is that disrupted REM sleep induces maladaptive changes in the medial (M) and lateral habenula (LHb), which in turn contribute to escalated alcohol intake and promote relapse. The key findings are: 1. Specific REM sleep deficits at 2-3 weeks after alcohol withdrawal are a robust predicting factor for relapse in alcohol dependent patients; 2. Chronic, selective REM sleep fragmentation enhances medial Hb neuron tonic firing and negative affect; 3. Relapse to alcohol seeking is associated with increased activation of MHb and LHb neurons. 4. Sleep fragmentation precipitates drug craving during abstinence from cocaine, while selective consolidation of REM sleep reduces craving. Thus, it is predicted that chronic alcohol drinking and abstinence leads to disrupted REM sleep, which induces Hb hyperactivity, promoting a negative affective state that drives motivation for alcohol and relapse. Likewise, it is predicted that chronic REM sleep disruptions can initiate this cycle, leading to increased risk for developing AUDs. Therefore, aim 1 in this proposal will evaluate the effects of chronic and escalating alcohol drinking and abstinence on sleep architecture and Hb neural activity. The proposed experiments will use the intermittent access two-bottle choice ethanol drinking paradigm combined with chronic EEG/EMG recordings and ex vivo slice electrophysiology. It is expected that the baseline REM sleep features or the REM sleep deficits following alcohol drinking/abstinence may predict the amount of alcohol drinking as well as Hb neural activity. Aim 2 will determine if chronic REM sleep fragmentation promotes the escalation of alcohol intake in the intermittent access model and/or promotes relapse in an operant self-administration model. We will further determine if the REM sleep effects on alcohol intake/relapse are, in part, mediated by REM sleep fragmentation-induced hyperactivity of Hb neurons. Aim 3 will test the treatment potential of interventions that improve REM sleep. Utilizing recently developed REM sleep-selective manipulations, it will be determined if selective REM sleep consolidation can reduce alcohol drinking and relapse, and if this is associated with reduced Hb activity. Overall, the proposed studies will provide a critical assessment of the potential for developing REM sleep- focused therapeutics for preventing relapse, and provide novel mechanistic insight into the interactions between alcohol intake and sleep homeostasis in the context of habe...