Cancer is one of the diseases for which RNA interference is a potential therapeutic approach. Genes involved in the promotion or maintenance of tumor growth are obvious targets for RNAi. RNAi is also considered an attractive additional approach to conventional chemotherapy for cancer treatment. Moreover, siRNAs have shown a high specificity for their molecular target mRNAs as they can selectively inhibit cancer-promoting genes that differ by a point mutation. Loss of heterozygosity (LOH) reduces genes to hemizygosity in cancer cells and presents an absolute difference between normal and cancer cells. The regions of LOH are usually much larger than the tumor suppressor gene, which is lost, and has been shown to contain genes that are essential for cell survival. Single-nucleotide polymorphisms (SNPs) are the most common type of genetic variation in man. SNPs in essential genes that are frequently affected by LOH can be used as a target for a therapy against cancer cells with LOH. We have designed siRNAs against the gene of the large subunit of RNA polymerase II (POLR2A), a gene located in close proximity to the tumor suppressor gene p53, which frequently shows LOH in cancer cells. It is shown in vitro that siRNA can selectively inhibit POLR2A expression dependent on its genotype. Furthermore, cancer cell proliferation and tumor growth inhibition in nude mice was genotype dependent. We conclude that siRNA can be used for genotype-specific inhibition of tumor growth targeting an SNP in POLR2A in vivo.