A growing body of evidence supports the hypothesis that the retinoic acid receptor beta2 (RAR-beta2) gene is a tumor suppressor gene which induces apoptosis and that the chemopreventive and therapeutic effects of retinoids are due to induction of RAR-beta2. During breast cancer progression, RAR-beta2 is reduced or even lost. It is known from studies of other tumor-suppressor genes that methylation of the 5'-region is the cause of loss of expression. Several groups demonstrated that this is also true for the RAR-beta2 in breast cancer by treating breast cancer cell lines with a demethylating agent and examining expression of the RAR-beta2 gene in response to a challenge with retinoic acid. Studies using sodium bisulfite genomic sequencing as well as methylation specific PCR showed that a number of breast cancer cell lines as well as breast cancer tissue showed signs of methylation. The RAR-beta2 gene was unmethylated in non-neoplastic breast tissue as well as in other normal tissues. A combination of retinoic acid with demethylating agents as well as with histone deacetylase inhibitors acts synergistically to inhibit growth. This review presents data that suggest that treatment of cancer patients with demethylating agents followed by retinoic acid may offer a new therapeutic modality. Both the time of commencement of chemoprevention and the choice of substances that are able either to prevent de novo methylation or to reverse methylation-caused gene silencing may be important considerations.