This study used dietary exposure assessment method and K-means++ clustering algorithm to construct a carbofuran risk grading model to assess the risk of carbofuran in six vegetable categories in 20 provinces in China using carbofuran sampling and testing data and consumption data in 2020. The clustering algorithm classified the risk level of carbofuran in vegetables into 3 levels: low, medium, and high. The number of low risk combinations accounted for 92.5%, and the high-risk combinations were bulb vegetables in Hebei and leafy vegetables in Shaanxi. This study uses objective data to build a risk classification model and a data-driven approach to risk classification, enhancing the objectivity and validity of the experimental results.