The melanoma staging system proposed by the American Joint Committee on Cancer (AJCC) (which classifies melanoma patients into four clinical stages) is currently the most widely used tool for melanoma prognostication, and clinical management decision making by clinicians. However, multiple studies have shown that melanomas within specific AJCC Stages can exhibit varying progression and clinical outcomes. Thus, additional information, such as that provided by biomarkers is needed to assist in identifying the patients at risk of disease progression. Having previously found six independent prognostic biomarkers in melanoma, including BRAF, MMP2, p27, Dicer, Fbw7 and Tip60, our group has gone on to investigate if these markers are useful in risk stratification of melanoma patients in individual AJCC stages. First, we performed Kaplan-Meier survival and Cox proportional multivariate analyses comparing prognostication power of these markers in 254 melanoma patients for whom the expression levels were known, identifying the best performing markers as candidates for stage-specific melanoma markers. We then verified the results by incorporating an additional independent cohort (87 patients) and in a combined cohort (341 patients). Our data indicate that BRAF and MMP2 are optimal prognostic biomarkers for AJCC Stages I and II, respectively (P = 0.010, 0.000, Log-rank test); whereas p27 emerged as a good marker for AJCC Stages III/IV (0.018, 0.046, respectively, log-rank test). Thus, our study has identified stage-specific biomarkers in melanoma, a finding which may assist clinicians in designing improved personalized therapeutic modalities.