Principal component analysis and cluster analysis of 19 plum germplasm resources which planted in Sichuan were conducted using 25 fruit characters and 19 SSR markers. The plum germplasm were classified based on Euclidean distance, the first three principal components and genetic similarity, as well as systematic cluster, three-dimensional ordination plot and UPGMA clustering analysis. Results showed that the average coefficient of variation of the 25 fruit characters was 54.71%, the variation coefficient of a* value of the peel was the highest (436.14%) and the variation coefficient of stone lateral dinmeter was the lowest (9.99%). The system clustering mainly divided these varieties into large fruit with deep color, small fruit with light color. In the principal component analysis, the first three principal components account for 73.27% of variation among the varieties, the first principal component (PC1) which regarded as fruit weight factor contributed 49.45%, the second principal component (PC2) as fruit shape and the color factor contributed 14.21%, and the third principal component (PC3) as nutritional factor contributed 9.62%. The three-dimensional ordination plot with PC1, PC2 and PC3 vectors clearly showed the relative differences of the main characters for the 19 plum germplasm resources. SSR markers showed that the genetic diversity of the tested materials was rich. The results of UPGMA clustering were similar to those of systematic clustering and three-dimensional ordination plot.