The prognosis of lung cancer is very much limited by the difficulties of diagnosing early stage disease amenable to surgery. Thus, novel diagnostic and therapeutic approaches are urgently needed for this common type of cancer. Recently, epigenetic alterations of tumor cells have been defined for a multitude of tissues and genes. Thus, promoter hypermethylation of tumor suppressor genes, and other targets of neoplasia-associated methylation disturbances, have become the most frequent recurrent alteration in solid tumors and hematologic neoplasia. In lung cancer, several sets of genes including the tumor suppressor gene p16, the DNA repair gene O(6)-methylguanine-DNA methyltransferase (MGMT), E-cadherin and retinoic acid receptor beta have been shown to be frequently methylated and inactivated. Distinct methylation patterns can provide molecular distinctions between different histologic subtypes of lung cancer. Gene hypermethylation in lung cancer is an early event associated with exposure to tobacco-specific carcinogens. Highly sensitive detection of hypermethylated DNA in sputum and peripheral blood offers a powerful tool for detecting lung cancer at an early stage. Epigenetic alterations in cancer, as opposed to genetic lesions, are potentially reversible. Thus, hypermethylation has been studied as a therapeutic target for agents which revert this epigenotype. The most advanced drugs to inhibit methylation are two azanucleosides, decitabine and its ribonucleoside analogue 5-azacytidine. In vitro, demethylating agents given at low doses reactivate tumor suppressor genes, and in mouse models, the development of lung cancer can be retarded. This effect is more powerful when histone acetylation, as a second epigenetic silencing mechanism, is also inhibited pharmacologically (HDAC inhibitors). Clinical trials of both groups of agents have been performed, and novel demethylating agents which are not incorporated into DNA offer further perspectives for epigenetic therapy of lung cancer and other malignancies.