Considering the global impacts of drug addiction, this article focuses on opioids, alcohol, and cannabis, emphasizing their influence on the sample population. It combines psychological assessment with predictive analysis using machine learning to understand and predict substance consumption tendencies. The study is designed to collect information from the participants using standard psychological assessment tools. An app-based survey is conducted to prepare a dataset based on consumption patterns of drugs and alcohol. Further impulsivity is also measured for the sample population (n = 1000) with 58% male participants. An associative importance vector is devised in consultation with a clinical psychiatrist. Every behavioral and consumption questionnaire factor is augmented to derive a final inclination score. Machine learning algorithms are applied to classify the training and testing dataset. The result observed that the random forest algorithm performed better than the others with an accuracy of 93%. The research aims to develop models predicting an individual’s likelihood of substance engagement by integrating psychological, demographic, and behavioral variables. The usage of such research includes early identification of at-risk individuals and tailored interventions, potentially revolutionizing substance abuse prevention. It further assists the medical fraternity in contributing to the intersection of psychology, machine learning, and substance consumption tendencies for improved public health outcomes.