This paper aims to understand the factors such as demographics and social determinants of health (SDoH) that influence treatment completion compared to dropout among young adults with substance use disorders (SUD). The study sample included 2909 young adults ages 18-25 enrolled in publicly-funded outpatient substance use treatment programs who also completed the Adult Needs and Strength Assessment (ANSA), a questionnaire designed to assess aspects of functioning of adults with mental health and substance use disorders. Participants’ Total Actionable Item (TAI) improvement rates were calculated as a measure of progress in treatment (equal to the number of actionable ANSA needs items at final assessment by the total number actionable needs items ever identified for an individual). Additionally, the National Outcome Measures (NOMS) tool was administered to participants when admitted to the treatment programs and repeated every six months until discharge. Initial descriptive statistics were calculated to investigate any differences between those who completed treatment and those who dropped out. Then, the Chi-square Automatic Interaction Detection (CHAID) approach was used as a machine learning (ML)-based decision tree model to identify various factors that may influence whether an individual completes treatment. This model was used in this study to determine combinations of characteristics such as demographics and SUD and treatment-related traits that distinguish people who completed treatment from those who dropped out. The CHAID analysis showed that TAI improvement rates were the strongest predictor of treatment completion. Treatment completion was most likely for those living in urban areas with family-related strengths and higher TAI scores. Comparatively low treatment completion rates were found for those with lower TAI scores, opioid use disorder, and justice system involvement. When the performance of the CHAID classification algorithm was evaluated (via area under the ROC curve), it received a score above that considered acceptable. The findings of this study indicate that there are multiple intersecting factors that influence completion of SUD treatment. Conclusively, the study found that treatment completion depended greatly on a measure of how fast symptoms are improving (TAI improvement rates), underscoring the importance of monitoring and promoting individuals’ treatment progress. Key differences were found between those in rural and urban communities, indicating the need to strengthen treatment resources in rural communities. Criminal justice system involvement was found to be associated with treatment completion, which may relate to legal requirements of those in the justice system and may speak to the punitive nature of substance use policies in the US. In conclusion, multiple demographic factors interact with treatment progress (TAI improvement rates) to predict SUD treatment completion versus dropout.