The hypermethylome detected in melanomas that represents better prognoses markedly decreased. The decrease in the methylation levels occurred gradually, as the continuous Breslow thickness variables allowed us to distinguish more than two groups among primary melanomas and to map the progress of demethylation during distinct stages. The genes involved in demethylation partially overlap among clinical subgroups: five genes were found to be commonly demethylated in large, nodular subtype, ulcerated and metastatic melanomas. The SEPT9 gene is an ovarian tumour suppressor playing a role in cell cycle control ; IL8 gene expression is elevated in metastatic melanomas and can increase the level of MMP2 ; SLC22A18 has been reported to be down-regulated due to promoter hypermethylation in gliomas ; MMP14 has not been found to play a role in melanoma progression thus far. Among the aforementioned clinical groups, the largest similarity has been detected between the demethylated genes associated with Breslow thickness and ulceration. The histologic subtype represents the most unique methylation pattern, comprising 30 differentially methylated genes between superficial and nodular melanomas. Our results contrast those of studies describing the hypermethylation patterns of specific genes as tumour progression-related markers based on single gene approaches. However, Conway et al. supported the claim that a covalent change from cytosine to 5methylcytosine in the promoter region occurs as an early aberration event in melanomas. Notwithstanding, their results highlighted not only the hypermethylated but also the demethylated genes in heterogeneous melanomas compared to naevi. This group reported a lack of similarity – involving only two genes, namely, RUNX3 and SYK – with the previously published data. In addition to the common mutations, specific patterns of CN alterations have been reported in melanomas characteristic of unfavourable clinical outcomes. Furthermore, it has become obvious that BRAFV600E mutated melanomas display distinct patterns for CN changes, providing the first line of evidence in support of Knudson’s two-hit hypothesis. However, none of the published studies attempted to evaluate the relationship between CN alterations and DNA methylation in melanomas. Our group performed a Tiling Array CGH, and, apart from highlighting common CN losses and amplification in the subgroups of primary melanomas, we demonstrated that 6q12– 6q25.1 comprises a remarkable CN loss, harbouring two hypermethylated genes on 6q23, EYA4 and MYB1. This result was measured and verified quantitatively and provides evidence for Knudson’s two-hit hypothesis at the level of CN loss and DNA hypermethylation. Notably, MYB1 is an important discriminator between melanomas and naevi, as validated by FISH in 123 melanomas and 110 naevi. The copy number deletion of MYB1 is currently used in the diagnosis of melanoma.