Pathological gambling is an addictive disorder characterized by repeated problematic gambling despite severe negative consequences, often linked to poor decision making and impulsivity. Research into gambling disorders has been facilitated by using rodent tasks that are directly translated from tasks used in humans. Traditional analyses for these tasks result in simplistic measures linked to decision making deficits and impulsive behavior. More recently in clinical research, computational modeling has been applied to data from human tasks to allow more sophisticated, in-depth analysis of the psychological and cognitive processes underlying these behaviors. Despite many parallels in task measures and structure, computational approaches are much less frequently used in the analysis of rodent behavioral data. Using multiple large datasets from rodent gambling and decision making tasks, this project proposes to use computational modeling to investigate the underlying cognitive processes involved in performance on these tasks. This will improve translatability of findings from preclinical research, and enable the generation of testable hypotheses for novel therapeutic/behavioral interventions to address problematic gambling behavior.