4522 Background: The objective of this study is to create a molecular tool that can be applied widely to clinical specimens using existing transcript signatures for use in clinical risk prediction of clear cell Renal Cell Carcinoma (ccRCC) to improve personalized disease management. Methods: We developed a 34-gene subtype predictor to classify clear cell tumors according to two subtypes, clear cell A (ccA) or B (ccB). The training set consisted of 72 ccRCC microarray-analyzed tumor samples that had previously been classified by unsupervised clustering and logical analysis of data (LAD). The predictor was developed from a panel of genes significantly expressed in ccA and ccB tumors and associated with prognosis. The prognostic value of the algorithm was corroborated in RNA-sequencing data from 379 ccRCC samples from The Cancer Genome Atlas (TCGA) and further validated using the NanoString platform with a cohort of 163 archival fixed samples collected at the University of North Carolina. Results: Risk associated molecular subtypes, ccA and ccB, were classified in TCGA and NanoString cohorts. Subtype classification showed significant prognostic outcomes for overall survival (p<.001), cancer-specific survival (p=.003), and recurrence-free survival (p<.05) and remained significant in multivariate analyses that included age at diagnosis, gender, ethnicity, pathologic stage, and histologic grade. A prognostic model was built for overall and recurrence-free survival for non-metastatic ccRCC patients within the context of subtype and clinical characteristics. Conclusions: The ccA and ccB subtypes significantly added prognostic information to clinical parameters, particularly for non-metastatic ccRCC patients.The subtypes can be used for future analyses involving risk for developing metastatic disease and cancer-specific outcomes. This research was supported with a grant from the American Association for Cancer Research, and the UNC Lineberger Comprehensive Cancer Center Cancer Cell Biology Training Grant.