Abstract
It is widely accepted that diet is important in the aetiology of cancer of the human gastrointestinal tract, but the rates of progress in unravelling the associations have varied at the different subsites, and with the different histological types of cancers. We have reached the stage of recognizing the problems, but have only begun to address them. Progress has been most rapid in the large bowel and least rapid in adenocarcinoma of the oesophagus and gastric cardia.