<p>The depletion of natural river sand and the accumulation of textile waste pose serious environmental challenges for the construction sector. The study investigates the feasibility of incorporating textile waste as a partial replacement for fine aggregate in concrete and predicts its mechanical properties using advanced ML models. A dataset comprising variations in textile waste content, cement dosage, and water–binder ratio was analyzed, with compressive, split tensile, and flexural strengths as outputs. Five ML models were developed and validated using cross-validation and multiple error metrics. Among these, CatBoost and Gradient Boosting achieved the highest predictive accuracy, while XGBoost showed overfitting tendencies. SHAP analysis identified cement content and water–binder ratio as dominant predictors, with textile waste influencing tensile and flexural performance through its fibrous morphology. Unlike previous ML-based studies that focused on conventional or single waste materials, the present research integrates a comprehensive literature-derived dataset of textile waste concrete and provides the first interpretable predictive framework for its strength behaviour. The findings confirm that moderate levels of textile waste can be incorporated into concrete without significant loss of strength, demonstrating both environmental and engineering benefits. This framework contributes to the circular economy by enabling predictive optimization of waste-based concrete mixes.</p>