Background: AJCC v8 includes Breslow thickness and ulceration to subdivide stage I and II CM pts into risk groups. In light of the results from adjuvant therapy in stage II CM, it has been discussed that pts’ follow-up and eventually treatment should consider additional markers, namely CP-GEP, to further refine the risk classification provided by the AJCC v8. The aim of this single center study was to clinically validate a prognostic CP-GEP-based risk score for stage I/II CMs combining Breslow, age and the expression of 8 genes SERPINE2, GDF15, ITGB3, CXCL8, LOXL4,TGFBR1, PLAT and MLANA. Methods: All obtainable formalin-fixed paraffin-embedded primaries of stage I/II CMs with negative sentinel lymph node (SLN) from the Central Malignant Melanoma Registry of Germany diagnosed between 2000-2017 and archived in Tuebingen were included. Study hypothesis and protocol were prospectively formulated. Tumors were analyzed blinded to clinical outcome. Quantitative reverse transcription polymerase chain reaction of the 8 genes was performed and combined with age and tumor thickness to define CP-GEP low- vs. high-score groups. Relapse-free survival (RFS), distant metastasis free survival (DMFS) and overall survival (OS) were evaluated using Kaplan-Meier curves. CP-GEP score performance was tested using multivariate Cox regression adjusted for tumor thickness, ulceration and age. Results: We included 543 pts with Stage IA (n=78); IB (n=223); IIA (n= 123); IIB (n=73); IIC (n=46). 43% were females, median Breslow was 1.7mm and 25% of tumors had ulceration. The median follow-up was 78 months (IQR 47-116). 311 (57%) patients had a high-risk CP-GEP score. The 5-y RFS rate was 71% and 92% (HR 4.2; p<0.001), the 5-y DMFS rate was 86% and 96% (HR 4.35; p<0.001) and the 5-y OS was 85% and 95% (HR 3.2; p=0.001), respectively for high and low-risk CP-GEP score. In multivariate Cox regression analysis for RFS including Breslow thickness, ulceration and age, contribution of CP-GEP score remained independently significant (HR 2.75; p=0.0008) compared to age (HR 1.03; p<0.0007), Breslow (HR 1.21; p<0.0001) and ulceration (HR 1.37; p=0.1694). Conclusions: CP-GEP risk score is a non-invasive and independent prognostic model for risk of relapse in stage I/II melanoma validated in this study. It identifies SLN negative pts at high risk of relapse and should be considered for complementing AJCC classification and for inclusion in future clinical trials.