The identification of potential human health hazards stems from the obvious need to prevent, avoid, reduce, and eliminate exposure to hazardous agents, mixtures of agents, or exposure circumstances. The first step in the risk assessment process centers on determining if a hazard exists. The following strategies are used for this purpose: (i) epidemiologic investigations, (ii) long-term chemical toxicology and carcinogenesis studies on laboratory animals, (ii) shorter-term in vivo and in vitro assays, and (iv) physicochemical structure-activity relationships. Indicator 1 is the most relevant and reliable if adequate data are available; indicator 2 is the most valid and useful alternative for human experience; indicator 3 allows certain toxicologic end points to be identified, but generally needs confirmatory and supportive information; and indicator 4 has made gains in the area of predictivity. The advantages and limitations of each are given. The magnitude of the overall cancer hazard identification effort and a likelihood number of eventual chemical carcinogens have also been estimated.