From 1972 to 1979, a total of 3392 patients with endoscopically proven ulcers were seen, in six ethnic groups. The distribution was as follows: Africans 456 males, 182 females; Muslim Gujerati Indians 206 males, 60 females; Hindu Hindi 433 males; total North Indians 639 males, 195 females; Hindu Tamils 593 males, 184 females; Hindu Telegu 179 males, 46 females; total South Indians 872 males, 230 females, and Whites 465 males, 303 females. In the continent of India, it is predominantly the South Indians who suffer from duodenal ulcer. In Durban, the number of North Indians with duodenal ulcers approximates that of those from the South (North: South ratio 0.83). The first question raised by this study is that the protective factors in North Indians in India are not genetic, and are lost when they emigrate to Natal. This may be due to changes in diet. A seasonal analysis indicates that, for females, there is a striking Autumn and Winter predominance in all Indian groups, reaching 80% in Muslims and Telegus but not in African females (52.7%). The second question raised by this study is that protective factors must be sought which operate in Indian females in the Spring and Summer months. The third question emanating from this study is that duodenal ulcers (and ischaemic heart disease) appear to increase in times of dietary and social change. This occurred in the West from 1890 to 1960, and is still occurring in the Third World. The restoration of dietary fibre and unsaturated fat, and the possible adjustment to stress in the West since 1960, has been accompanied by a fall in the incidence of these diseases. A 'changing factors' theory of duodenal ulcers and ischaemic heart disease is proposed. These conditions fall when a 'plateau situation' is reached.