For each kind of occupationally associated cancer, there are three distinct stages in the development of the problem: recognition of a possible problem, confirmation, and the introduction of preventive measures. In the past, recognition of a possible problem depended heavily on chance and on the powers and observation of dedicated physicians and surgeons. Confirmation consisted of the collection of further anecdotal evidence and the conduct of case-referent (case-control) or other studies. The introduction of preventive measures often lagged woefully behind confirmation that a problem existed. Recently, the power of epidemiology as a primary investigative tool has grown to the point where unsuspected associations between occupation and cancer risk may be the first hint that a problem exists. However, it is important to recognize that investigative epidemiology is capable of constructing misleading pictures. In the future there is bound to be continuing pressure to reduce maximal permissible exposure to proven carcinogens. For chemicals for which there is no more than suspicion based on laboratory tests, one must ensure that regulatory action is based on good science, sound judgement, and common sense, rather than on the machinations of those with vested interests, of ambitious lawyers, or of the lunatic fringe. Less than 10% of all cancers are likely to be due to occupational factors. Therefore, even turning the world upside down with safety precautions against actual and suspected carcinogens would only marginally affect the present human cancer burden.