Scientific advances in recent years have tremendously improved the predictive capabilities of domain-specific problems with the use of machine learning and artificial intelligence. An innovative exploration is performed to understand the root uptake of per- and polyfluoroalkyl substances (PFASs) by plants, focusing on the intricate interactions between PFAS compounds, crops and soil. We established a machine learning model which performs a regression task to accurately predict the root concentration factors (RCFs) values of the PFAs. Various machine learning models are trained and evaluated on various evaluation metrics, and the best model has an R^2 value of 0.9379. These models significantly outperformed the other existing models in predicting the logRCF values, indicating their robustness in capturing the complex dynamics of PFAS uptake by plants. For model development, around 300 instances (or data points) of root concentration factors (RCFs) that measure the amount of PFASs absorbed by the plant roots from the soil are used. The data also included 11 features, related to PFAS chemical structures, organic carbon content, crop and soil characteristics and cultivation conditions. The developed model is evaluated and interpreted to obtain the most important features, which highly contribute to predicting RCF values. Feature importance analysis was utilized to gain a greater understanding of the decision-making processes of the models and the significance of individual features. This study shed light on a detailed approach to predict and understand how plants absorb PFASs and also captured the essential variables that affect the uptake of PFAS, and offered insightful information about the various components that contribute to their occurrence.