The process of tumor progression was described by Rous in 1939 as that by which "tumors go from bad to worse".In recent years, this has been attributed to the sequential acquisition of successive genetic abnormalities, including the activation of oncogenes and the loss of suppressors.Epigenetic mechanisms are also involved in these processes,including methylation changes, not well studied so far in melanoma, and micro RNA genetic suppression.The clinical morphology of the lesions of progression in melanoma has been well described, including definitions of benign melanocytic tumors including nevi and dysplastic nevi, the early curable phase of melanoma termed radial growth phase, the more advanced primary melanomas termed vertical growth phase or tumorigenic primary melanomas, and metastatic melanomas which are more often than not incurable when spread has occurred beyond the region.This clinical morphology presently serves as the gold standard for diagnosis of melanoma and therefore provides the primary baseline for evaluation of diagnostic marker studies.Although there are no markers that reliably distinguish between benign and malignant lesions, proliferation-associated molecules such as Ki-67 have some value for this purpose.In addition, melanoma lineage-specific markers such as MelanA/Mart and MITF are used to highlight diagnostically important architectural patterns.The cell cycle suppressor molecule p1 6 may be of some value in distinguishing between Spitz tumors and morphologically similar tumorigenic melanomas in children.Reliable markers of prognosis in melanoma include ulceration, mitotic rate, tumor-infiltrating lymphocytes, gender, site, and regression, features that are not ordinarily considered to be "markers".Ki-67 may have some value in prognosis but adds little to the information provided by mitotic rate.The most important predictive markers now in use involve the analysis of activating mutations in oncogenes.Oncogenes for which effective targeted therapy are available include BRAF and C-kit.Other oncogenes have been described for which targeted therapy will likely be developed.At present, most tumors escape from their susceptibility to the presently available agents and it is likely that multiple agents will need to be used synergistically.There is considerable promise for a future in which metastatic melanoma may be controllable with targeted agents.For the present, the most reliable means of reducing long-term mortality from melanoma remain the use of public education to reduce exposure to etiological agents such as sunlight, and clinical skills to enable diagnosis of melanoma in its early, curable stages.