Arsenic in certain forms and amounts can kill after a single exposure—this is the basis for the association of arsenic with poison. But it is chronic (repeated) exposure to smaller amounts of arsenic that poses the greatest concern. We know most about the toxicity of inorganic arsenic (those arsenic compounds that do not contain carbon) but, despite many decades of intense study, it is not completely clear how arsenic causes disease. We do know that arsenic leaves no organ in our body untouched, and the list of potential ailments is long (skin lesions, nervous system disorders, cardiovascular disease, diabetes and several types of cancer). It is also apparent that exposure in early life (in utero or in childhood) can be particularly harmful and that some effects may take decades to reveal themselves. The likelihood of developing an arsenic-related disorder depends on the extent of exposure. As zero exposure is not possible, it would be good to know whether there is a safe level of exposure or at least one that poses an “acceptable” risk. In this chapter, we outline why there is not a clear answer to this question and describe how risk assessment is used to minimize risk.