A total of 21 different regions were found to have a very low relative frequency and/or low incidence rates of esophageal cancer (the male mean being 0.6 and the range being 0.1--1.0/100,000 per annum). In all these areas the dietary staples were either sorghum, millet, cassava, yams, or peanuts or a combination of these items. In another 17 areas that had a high risk for esophageal cancer (the male mean being 41.3 ad the range being 16.5--86.0/100,000 per annum), the dietary staple was invariably corn or wheat. It is calculated that dietary staples associated with a high risk for esophageal cancer will be marginal or deficient particularly in riboflavin, nicotinic acid, magnesium, and zinc, whereas dietary staples associated with a low risk for esophageal cancer will be rick in these substances. the evidence presented supports the concept that these high rates of esophageal cancer in diverse peoples are associated with long-standing deficiencies of a few micronutrients and explains epidemiologic features such as geographic variation, recent emergence of the disease in Africa, and the role of alcohol abuse.