PROJECT SUMMARY To make good decisions we mentally cast ourselves into the future to consider the identity and value of the outcomes of our potential choices. This prospective consideration is facilitated by an encoded internal model (aka cognitive map) of environmental associative relationships, e.g., between stimuli and the specific rewards they predict. Such cue-reward learning enables one to contemplate potential forthcoming events and facilitates adaptive decision making. While we know of brain regions that contribute, the pathways and circuits of cue-reward learning and decision making are less explored. Given that substance use disorders are characterized by deficits in cue- reward learning and decision making, this gap in knowledge limits clinical advance. The goal of this project is, thus, to expose the neuronal pathways and circuit motifs that support the cue-reward learning and memory processes that enable adaptive decision making. In the last funding period, we discovered the basolateral amygdala (BLA) is a hub for the encoding and use of the detailed, identity-specific cue-reward memories, i.e., cognitive map, that enable flexible decision making. The BLA output pathways that subserve these functions are largely unknown. Our preliminary data suggest the BLA may mediate cue-reward memory via projections to the lateral (lOFC) and medial (mOFC) orbitofrontal cortex. Activity in these projections is critical for using cue-reward memories to guide decision making, with the BLAlOFC and BLAmOFC pathway needed for identity and value predictions, respectively. The information conveyed by these pathways and whether they also support cue-reward learning is unknown. The BLA may also mediate cue- reward learning and decision making via projections to the striatum, including the dorsomedial striatum (DMS), nucleus accumbens (NAc) core and shell. Whereas DMS and NAc shell have been implicated in choices based on predicted reward identity, NAc core has been implicated decisions based on predicted value. Little is known about how these functions are achieved. Our general working hypothesis is that BLA projections to lOFC, NAc shell, and DMS contribute to encoding and using identity-specific cue-reward memories to enable adaptive decisions, whereas BLA projections to mOFC and NAc core contribute to aspects of cue-reward memory that enable decisions to adapt based on predicted reward value. We will test this hypothesis using a suite of modern systems neuroscience tools including pathway-specific, optogenetic neuronal activity monitoring and bidirectional manipulation coupled with precision behavioral assessments of cue-reward learning and its influence on decision making with outcome-specific Pavlovian-to-instrumental transfer and devaluation tests. We will expose critical functions of 5 BLA output pathways, fill important gaps in knowledge of the neuronal pathways that support cue- reward learning and decision making, and reveal an appetitive learn...