Metals are an important and emerging class of carcinogens although their mechanism of action is largely unknown. For cadmium (Cd) we have found clear strain differences in carcinogenic sensitivity, and a new target site, the hematopoietic system, was detected in mice. Cd's potential as a prostatic carcinogen has been studied and several chronic experiments show it can produce tumors of the rat prostate. Most recently we have found that the metallothionein (MT) gene produces a protein clearly associated with tolerance to Cd genotoxicity and is poorly expressed and unresponsive to stimulation in the specific lobe of the rat prostate sensitive to Cd carcinogenicity. Likewise, the testicular MT gene is not activated by metals and does not play a role in the induced tolerance to Cd carcinogenesis generated by zinc (Zn). Thus, there is evidence that the MT gene plays a fundamental role in sensitivity to the carcinogenic and genotoxic effects of Cd. Our studies also reveal the serendipitous finding that Cd can also be a very effective tumor suppressant. Most recently we studied suppression of N-nitrosodiethylamine (NDEA)-initiated tumors in mice by a single i.v. dose of Cd given well after liver tumor formation. The Cd dramatically reduced (greater than 50%) the hepatic tumor incidence. Multiplicity and size of liver tumors were also markedly diminished by Cd. Clear evidence of tumor specific toxicity was seen as Cd-induced necrosis in the hepatic tumors. MT was not detected in liver tumors, even those undergoing Cd-induced necrosis, while normal liver cells expressed high levels. The possibility exists that Cd has a specific toxicity toward initiated cells within liver and lung because the MT gene is only poorly expressed in such cells. These results indicate that Cd can dramatically reduce tumor incidence in certain organs by tumor cell specific cytotoxicity. This chemotherapeutic potential of Cd deserves further study.