Background: A GEP has been validated as an independent predictor of DM, distant metastasis free survival (DMFS) and overall survival (OS) (Gerami, CCR 2015; Gerami JAAD, 2015). This abstract combines the third validation cohort study with the two prior studies for analysis by T factor subgroups.Methods: 555 patients (pts) were enrolled in a multi-center IRB approved archival tissue study with a primary endpoint of DMFS. 492 CMs had T-factors available for analysis. Quantitative RT-PCR and Radial Basis Machine modeling classified CM tumors as low risk Class 1 vs. high risk Class 2. Results are reported for thin (T1), intermediate (T2/T3) and thick (T4) lesions. Most patients were node negative clinically or by sentinel node biopsy.Results: GEP Class 1 vs 2 was found to be a significant predictor of DMFS and OS (log-Rank p<0.0001), and GEP, Breslow’s thickness, and ulceration were independent predictors in Cox multivariate analysis (DMFS HR=2.7, p=0.0003; 2.2, p=0.03; 3.3, p<0.0001, respectively and OS HR=2.8, p=0.0002, 2.5, p=0.01, 1.7, p=0.01, respectively). In the T2/T3 N0 cohort (n=180) frequency of class 2 signature increases with depth [T2:35/78 (45%) Class 2; T3:75/102 (74%) Class 2] and ulceration [nonulcerated: 54/108 (50%) Class 2; ulcerated: 53/63 (84%) Class 2]. The table reflects survival analysis in this cohort. In the T1 group (median followup for non-DM group = 7.4 yrs) 14/173 (8%) reported DM. For T4 group: 42/80 (53%) developed DM. Data not considered adequate for further analysis of either group.Conclusions: GEP offers prognostic information that complements conventional staging at least in patients with T2/T3 CM. Ongoing studies will further define the role of GEP in evaluating these patients.