With the acceleration of global economic integration and increasing supply chain complexity, comprehensive capability assessment of third-party logistics service providers has become a core component in optimizing supply chain management. Addressing limitations of traditional evaluation methods—including high subjectivity, unreasonable indicator weighting, and difficulty handling nonlinear relationships—we propose a third-party logistics evaluation model based on the random forest algorithm. By integrating three key dimensions—functional indicators, operational metrics, and stability indicators—we establish a comprehensive evaluation framework and detail the specific application process and optimization strategies for the random forest model in logistics assessment.