At present, the available molecular markers of elm have been seriously deficient, which can not meet the needs of molecular marker assisted breeding and evaluation of germplasm resources. In this study, transcriptome data of Ulmus pumila were used as materials to develop and verify EST-SSR markers of elms. The availability of primers were checked in different elm resources. Moreover, the genetic diversity of different elm resources were analysised by using the developed primers. In this study, 8828 perfect SSR and 569 compound SSR locis were detected in 36609384bp sequences of Unigene sequences by using the transcriptome data of Ulmus pumila. The length of perfect SSR sequence were mainly based on short sequence of 10-22bp. The largest proportion of SSR repeat units were A/T(3330，40.18%), followed by AG/CT(1211，14.61%) and AAG/CTT(568，6.85%). In this study, we selected 90 pairs of EST-SSR primers randomly for validation analysis, the effective amplification rate was 51.11% (46 pairs) of which 63.04% primers (29) were high polymorphism primers. The polymorphism information content PIC were ranged in 0.054-0.683, which greatly enriched the resources of the elm SSR primers. This study confirms the availability of SSR primers in different elm resources. Cluster analysis showed that the majority of elm clones were separated from their origin, which proved the validity of the EST-SSR primers that developed in this study from the other side. This study provide the molecular basis for the classification of Ulmaceae plants.