Oesophageal cancer is common in Africa, and is a major cause of death in Transkei, South Africa. The cause of this endemic disease has been debated for many decades. A continuing focus of research has been identification of a single potent carcinogen. I argue that endemic incidence of oesophageal cancer is associated with potent initiation, not with potent carcinogenesis. Tobacco is a known oesophageal carcinogen, and there is very strong evidence that it is causally involved in a significant proportion of victims. Other potential carcinogens present in the environment include fungal mycotoxins, human papillomavirus, Solanum nigrum and nitrosamines. These are all of lower carcinogenic potential for the oesophagus, and do not have strong evidence associating them with the disease. In the presence of potent environmental initiation, any oesophageal carcinogen even if of low potency or of low concentration may cause the disease. Any or all of the substances named above may be involved, any one of them the cause of the final carcinogenic change in the individual. Tobacco exemplifies this point. It is of relatively low concentration/usage in Transkei, yet has an undeniable association with oesophageal cancer. Carcinogenesis for the oesophagus in Transkei is solely or predominantly due to agents which are already known, including tobacco, acting on a predisposed mucosa. The search for further carcinogens is of low importance, and the search for a single potent carcinogen is misguided. What is of importance is the development of methods to reduce the risks associated with predisposing factors and with known carcinogens.