Relapse is a central problem associated with the treatment of drug addiction. Exposure to brief stress can enhance craving and trigger relapse in humans, and triggers robust relapse to drug-seeking behavior in animal models. Recent work over the past ten years from our lab and others' has shown that a single exposure to an addictive drug can either initiate or block forms of synaptic plasticity in the ventral tegmental area (VTA), a region known to be critical in relapse to drug-seeking. We found that brief acute stress blocks synaptic plasticity at GABAergic synapses on VTA dopamine neurons (LTPGABA), and have characterized the molecular cascades involved. A single stressful experience activates kappa opioid receptors (κORs), and we showed that interrupting kappa receptor signaling rescues the LTP in brain slice recordings - but also importantly, prevents stress-induced reinstatement of drug-seeking. We have recently found that interrupting κOR signaling even days after the stressor both rescues LTP in dopamine neurons in vitro and prevents reinstatement of drug seeking in behavioral tests. While all the mechanistic steps are not yet known, all of our data support the general idea that removal of this normal brake on VTA dopamine neurons contributes to reinstatement. In this application, we will identify the circuits involved, asking: 1) which GABAergic afferents normally exhibit LTPGABA and lose it after acute stress, and 2) which dynorphin-releasing afferents and κORs are active during acute stress and alter the VTA circuit to drive relapse behavior. Knowing the brain regions and specific pathways involved offers the opportunity to manipulate these towards therapy development.