Trivalent and pentavalent forms of inorganic arsenic are ubiquitous elements found in nature that unfortunately result in significant human exposure. Oral exposure to arsenic occurs primarily from contamination of drinking water and food constituents, and is particularly high in certain regions of the world including areas of the southwestern United States, eastern Europe, India, China, Taiwan, and Mexico (1,2). Humans can also be exposed to arsenic through inhalation. This occurs primarily in occupations involved in mining/smelting operations, agriculture, or microelectronics (3,4). Epidemiological studies have demonstrated that exposure to inorganic arsenic is associated with increased risk of cancers of the skin and internal organs, including the urinary bladder, respiratory tract, liver, and kidney in populations from Finland, Taiwan, China, Bangladesh, Mexico, southwestern United States, and Central and South America (3-8). Arsenic-induced skin cancers usually develop 20 to 30 years after exposure, and occur in sun-exposed as well as nonexposed areas. The types of skin tumors found include either Bowen's disease, squamous cell carcinomas, basal cell carcinomas, or combined lesions (9-11). The key to identifying patients with arsenic-induced skin tumors is that they normally occur at multiple sites and unusual locations. Internal tumors are also common and are most frequently associated with the bladder. The association between arsenic exposure and urinary bladder cancers, typically transitional cell carcinomas, has been observed in the same endemic areas of the world where skin cancer populations have been identified. Lung tumors from arsenic are often associated with occupational exposure, such as smelters or agriculture workers, and occur from inhalation (12). In addition to neoplasia, additional pathological manifestations of chronic arsenic exposure include skin hyperpigmentation and hyperkeratosis (9,13), as well as vascular disease (14,15). Hyperpigmentation is the most common effect in individuals and can occur at any body site, and already pigmented areas are more accentuated. Arsenic-induced hyperpigmentation occurs almost exclusively in individuals of Oriental descent, although the genetic basis for this is not understood (16). Hyperkeratosis, which can appear within 4 years of exposure to arsenic, is manifested primarily in the form of hyperkeratotic papules or plaques and are most commonly found on the palms and soles (17). There are reports of cellular atypia at the base of these papules and on occasion their transformation into basal or squamous cell carcinoma. In contrast to carcinogenicity, little is known regarding the vascular effects of arsenic and most of the reports originate from individuals living in inner Mongolia, Xinjiiang, Toroku, and Nakajo (14,15,18). Circulatory manifestations of arseniasis include increased prevalence of ischemic heart disease and peripheral vascular disease. The latter is commonly known as blackfoot disease in southwestern Taiwan. In addition, dose-response relationships between arsenic exposure and hypertension prevalence have been reported in Southwestern Taiwan (19) and Bangladesh (20). On the basis of numerous epidemiological studies, arsenic has been classified as a potent human carcinogen, and population cancer risk due to arsenic has been suggested to be comparable to environmental tobacco smoke and radon in homes with risk estimates of around 1 per 1000 (11). It has been estimated that over 350,000 people in the United States consume drinking water containing over 50 ug/L of arsenic, the current EP A standard, and more than 2.5 million people use water containing more than 25 ug/L of arsenic (21). Subsequently, there is significant regulatory pressure to lower the acceptable levels. However, epidemiological studies, where exposure levels have been collected, suggest that the current EP A cancer slope factor (CSF) for arsenic may actually overpredict cancer cases at relatively low exposure levels (22). This may be due to the fact that the CSF was calculated assuming a standard linear dose-response relationship while a nonlinear or sublinear dose response may be more appropriate. Human epidemiological data are available providing empirical evidence supporting both a linear (23) and nonlinear (24) association between excess cancer and arsenic exposure. As will be discussed in the following sections, although the precise carcinogenic mechanism for arsenic has not been established, molecular, cellular, and metabolic studies suggest that a nonlinear relation may be most appropriate.