Cancer is nowadays recognised as a genetic and epigenetic disease. Much effort has been devoted in the last 30 years to the elucidation of the 'classical' oncogenes and tumour-suppressor genes involved in malignant cell transformation. However, since the acceptance that major disruption of DNA methylation, histone modification and chromatin compartments are a common hallmark of human cancer, epigenetics has come to the fore in cancer research. One piece is still missing from the story: are the epigenetic genes themselves driving forces on the road to tumorigenesis? We are in the early stages of finding the answer, and the data are beginning to appear: knockout mice defective in DNA methyltransferases, methyl-CpG-binding proteins and histone methyltransferases strongly affect the risk of cancer onset; somatic mutations, homozygous deletions and methylation-associated silencing of histone acetyltransferases, histone methyltransferases and chromatin remodelling factors are being found in human tumours; and the first cancer-prone families arising from germline mutations in epigenetic genes, such as hSNF5/INI1, have been described. Even more importantly, all these 'new' oncogenes and tumour-suppressor genes provide novel molecular targets for designed therapies, and the first DNA-demethylating agents and inhibitors of histone deacetylases are reaching the bedside of patients with haematological malignancies.