In the world of easy things truth is opposed to lie; in the world of complicated things one profound truth is opposed to another not less profound than the first. Neils Bohr The hypothesis of the exclusively genetic origin of cancer ("cancer is a disease of genes, a tumor without any damage to the genome does not exist") dominated in the oncology until recently. A considerable amount of data confirming this hypothesis was accumulated during the last quarter of the last century. It was demonstrated that the accumulation of damage of specific genes lies at the origin of a tumor and its following progression. The damage gives rise to structural changes in the respective proteins and, consequently, to inappropriate mitogenic stimulation of cells (activation of oncogenes) or to the inactivation of tumor suppressor genes that inhibit cell division, or to the combination of both (in most cases). According to an alternative (epigenetic) hypothesis that was extremely unpopular until recently, a tumor is caused not by a gene damage, but by an inappropriate function of genes ("cancer is a disease of gene regulation and differentiation"). However, recent studies led to the convergence of these hypotheses that initially seemed to be contradictory. It was established that both factors--genetic and epigenetic--lie at the origin of carcinogenesis. The relative contribution of each varies significantly in different human tumors. Suppressor genes and genes of repair are inactivated in tumors due to their damage or methylation of their promoters (in the latter case an "epimutation", an epigenetic equivalent of a mutation, occurs, producing the same functional consequences). It is becoming evident that not only the mutagens, but various factors influencing cell metabolism, notably methylation, should be considered as carcinogens.