<p>Non-cognitive skills are crucial for both educational success and personal fulfilment. Collectively, they constitute an essential skill set that enables students to excel academically and engage as active citizens. Although studies have explored the development of these critical skills, few have harnessed the potential of machine learning techniques. Accordingly, the current study aims to identify the most influential factors that affect non-cognitive skills using a broad range of determinants. Conventional Ordinary Least Squares (OLS) and supervised machine learning algorithms, including Least Absolute Shrinkage and Selection Operator (LASSO), Gradient Boosting Machine (GBM), and Random Forest (RF), were employed to evaluate the predictive performance using data drawn from the Programme for International Student Assessment 2022 (N=75,217). Findings indicate that variables across different groups contributed significantly to students’ non-cognitive skills, with cognitive-motivational traits emerging as the most influential predictor group. In terms of model performance, results demonstrate that advanced machine learning techniques outperformed conventional statistical analysis in identifying and ranking key predictors. Overall, this study highlights the complex, multifaceted nature of non-cognitive skills by identifying several variable-level predictors. Moreover, application of machine learning algorithms in this context represents an effective, novel analytical approach that can serve as foundation for future studies.</p>