Concrete in which plastic particles replace part of the fine aggregate is an innovative environmentally friendly concrete. Its main purpose is to reduce the dependence on natural aggregates in traditional concrete, while reducing environmental pollution by recycling waste plastics. This study develops a novel machine learning design and prediction of environmentally friendly concrete using plastic particles, which can drive sustainable improvement of building materials towards net-zero emissions. In this study, the compressive strength of the concrete with plastic particles is predicted using Random Forest (RF), XGBoost, and Artificial Neural Network (ANN) models. Initially, 132 pieces of experimental datasets from the public literature were collected, in which partially fine aggregate was replaced with plastic particles. The datasets were tuned by the grid search algorithm (GSA). The coefficient of determination (R2), mean square error (MSE), and mean absolute error (MAE) were employed to indicate the prediction ability of the machine learning models. The results demonstrate that the XGBoost model has promising abilities for design assistance in concrete with plastic particles.