Project Abstract Cocaine use disorder remains a significant public health problem in the US today, and there is a high risk of relapse even after long periods of abstinence. The current translational pipeline relies on animal models such as the extinction-reinstatement model to screen potential therapies for efficacy at attenuating relapse. While many pharmacological agents successfully reduce cocaine-seeking in this model, these agents show little clinical efficacy, and none have been FDA-approved for cocaine use disorder. One explanation for the failure of these agents to translate to the clinic may lie in the fact that most cocaine users engage in poly-substance use (PSU), while existing animal models of cocaine addiction involve self-administration of cocaine alone. Such PSU likely engages different behavioral and neural mechanisms compared to cocaine alone. Indeed, our preliminary data from a rat model of combined cocaine and alcohol use show that alcohol co-consumption significantly changes the neurobiology supporting cocaine relapse, and renders potential pharmacotherapies ineffective. These data and others highlight the need for a better understanding of PSU. Progress in this area is hampered, however, by a paucity of information regarding how substance (particularly cocaine) users actually engage in PSU. The long- term goal of this project is to determine the unique consequences of PSU on behavior and neurobiology underlying cocaine-seeking. The objectives of the current proposal, which represent the first steps toward our long-term goal, are to 1) develop and validate a survey instrument for evaluating detailed temporal patterns of PSU in cocaine users; 2) determine in a cocaine-using population the most common temporal patterns of alcohol and cannabis use (which are the most frequently used substances in combination with cocaine); 3) back- translate these data to develop rat models of cocaine+alcohol and cocaine+cannabis use; and 4) determine their consequences on neurobiological measures relevant for relapse (glutamate signaling and D2/3 dopamine receptor expression in the nucleus accumbens). Our rationale is that rat models which more closely mimic actual patterns of human substance use should better yield the underlying neuroadaptations present in humans, and should thus serve as better platforms for therapeutic discovery. As such, our central, unified hypothesis is that cocaine users will display high rates of comorbid cocaine+alcohol or cocaine+cannabis use in unique patterns that can be translated into rat models, in which the neurobiology underlying relapse to cocaine-seeking will be altered by such alcohol or cannabis use.