Despite reports of declining methamphetamine use in the early 2000’s, psychostimulant-related overdose deaths in the US, of which methamphetamine is primarily involved, increased ~1,800% from 1999-2017. Currently, identifying IDM who are in the greatest need for intervention is only discovered after catastrophic events have occurred (e.g., overdose, arrest, job loss). Thus, there is an urgent necessity for objective means of identifying IDM at-risk for such devastating consequences, before these consequences can occur. Previous studies have shown correlations between independent measures of biological, psychological, and social factors and critical outcomes (e.g., substance use patterns) in individuals dependent upon methamphetamine (IDM). However, no research has leveraged the combined power of biological, psychological, and social measures to the predict outcomes in IDM. Recent success combining neuroimaging (biological) and psychosocial measures with advanced machine-learning techniques to predict treatment or diagnostic outcomes in substance use and other disorders establish a precedent for achieving similar success in IDM. This project seeks to collect the first ever neuroimaging (magnetic resonance imaging [MRI], psychological (e.g., anxiety, depression, and psychosis symptoms), and social (e.g., interpersonal violence histories, peer network drug use) dataset from IDM (Aim 1). MRI data will be collected using state-of-the art sequences adopted from the Human Connectome Project and all data will made openly available allowing for the global scientific community opportunities to gain limitless insights from these unique data. Here, data will be used in machine-learning models for the prediction of two critical outcomes in IDM: substance use patterns and occupational functioning over a 6-month time period (Aim 2). Machine-learning models developed as part of this project will result in future funding proposals focusing on extending longitudinal analyses (e.g., 3-5 years) and acquiring additional participants for model evaluation. The long-term goal of this research is the development of predictive models quantifying an IDM’s probability of improving or worsening addiction and addiction-related consequences. In this way, this project will provide the foundation for objective, biopsychosocial models tailored toward the prediction and eventual prevention of greater harms to IDM.