The 5-year survival rates are estimated to be 50% for stage II, 20% for stage III and 5-10% for stage IV disease. Molecular genetic data supports the idea that endometrial carcinomas are likely to develop through a multi-step process of oncogene activation and tumor suppressor gene inactivation. In addition, studies have demonstrated that molecular alterations are specific for type I and type II endometrial carcinoma. Type I cancer is BAY 43-9006 characterized by mutation of PTEN, and defects in DNA mismatch repair . Type II cancers are characterized by p53 and Her-2/neu mutations. However, these gene alterations alone do not explain the different behavior and outcomes in type I and type II cancers. Most of the studies using cDNA microarray analysis have only focused on defining differential gene expression among different histologic types of endometrial cancer. The aims of this study were focused on stages and outcomes in two histologic types: EAC and USC. The purpose was to identify the difference in gene expression patterns in late stages versus early stages in each of the two types and in patients with good prognosis versus poor prognosis in EAC and USC, respectively. The main focus of the study is to identify and evaluate gene expression patterns in late vs. early stage and good vs. poor prognosis in each of endometrioid and serous types. Using the Illumina HumanHT-12 v3 microarray, we found 274 and 111 stage-related DEG in each of USC and EAC respectively. However, we were not able to find any overlap for DEGs with at least 2 fold changes in late stages EAC and USC versus early stages EAC and USC, indicating that tumor progression of different endometrial adenocarcinoma subtypes might be characterized by distinct gene expression signatures. In addition, we found 112 DEGs that are differently expressed for good vs. poor prognosis in EAC and 135 DEGs in USC. Only 1 DEGs is shared by the 135 prognosis-derived DEGs in USC and the 112 prognosis-derived DEGs in EAC. These findings indicate that tumor outcome of different endometrial adenocarcinoma subtypes might also characterized by distinct gene expression signatures. In other words, DEGs derived from USC might exclusively contain USC-specific prognosis biomarkers, while the DEGs derived from EAC might exclusively contain EAC-specific prognosis biomarkers. This confirms the existence of a distinct gene expression signature between endometrioid and serous carcinoma as previously seen, and that there is a distinct gene expression signature driving late vs. early stages in each of these two types. Additionally, the lack of overlap between stage-related and prognosis-related DEG sets in each subtype is of particular interest suggesting genes that drive and control stages might be distinct from the genes that drive and control outcomes. As a result, it would be of paramount interest to identify tailored biomarkers for outcome prediction and treatment modalities in patients with endometrial adenocarcinoma.