Compelling experimental and epidemiological evidence involves oxygen radicals in carcinogenesis, acting reactive oxygen species both as endogenous genotoxins during cell initiation and as messenger molecules in mitogenesis and in tumor promotion. Moreover, oxidants stimulate neoangiogenesis, which is a prerequisite for tumor growth. However, while several natural as well as synthetic antioxidant compounds appear to be chemopreventive in mutagenicity assays, antioxidant-based treatments for the prevention or cure of cancer have led to non-conclusive if not disappointing results. This is likely due to the fact that oxygen radicals have also a major role in the natural defences against the propagation of cancer cells, i.e. tumor cell apoptosis and immune surveillance, and mediate the beneficial cytotoxic effect of both the chemo-and radio-therapy. In recent years, the mitochondrial antioxidant enzyme, Manganous Superoxide Dismutase (MnSOD), has received a growing attention as a negative modulator of cellular apoptosis and as a survival factor for cancer cells. In fact, while overexpression of this enzyme in cancer cells decreases proliferation and tumor incidence in transgenic models, it is clear that even small amounts of this enzyme are crucial for cell resistance to inflammatory stimuli and anticancer drugs, and prevent oncogene-induced apoptosis triggered by the tumor suppressor protein p53. A previously unexpected oncogenic potential of MnSOD is also suggested by the elevated levels of this enzyme in several classes of human neoplasms, in a fashion which often correlates with the degree of their malignancy. This review focuses on the debated issue of the pro- and/or anti-tumoral effect of MnSOD, with special emphasis on recent observations suggesting that pharmacological inhibition of MnSOD may represent an effective strategy to selectively kill cancer cells and to circumvent their resistance to the commonly used anticancer treatments.