Port logistics supplier selection is a key link of global supply chain management in manufacturing enterprises. Decisionmaker’s perceived uncertainty affects supplier decision performance. By combining the feature of Structural Equation Modeling (SEM) and Data Envelopment Analysis (DEA), this paper proposes a SEM/DEA method for selecting port logistics supplier in multidimensions. Based on a large number of industrial sample data, this method is used to determine the weights of different indicators’ dimensions according to industrial preference, which can reduces the negative impact of single decisionmaker’s perceived bias to the result of selection and improves the rational degree of port logistics supplier select decision.