Abstract: In the past, we typically used Z score in the statistics analysis of the cotton-inspection-proficiency testing. But Z score has some limitations in the section of the reference number and analysis. Based on this problem, this paper will apply the grey clustering analysis which depends on the Trigonal Whitening Weight function to devide the cluster objects into better and general categories through taking the fourteen laboratories as the cluster objects and cotton’s upper half mean length of five cluster indicators. Through the theoretical and experimental validation in the statistical analysis of the testing, this method can be a statistical modal for the similar proficiency testing in the future.