The relationship between intake of dietary fibre and risk of colon cancer has been studied for 30 years and still the data are inconclusive. There are many possible reasons for this outcome, and they include a failure to consider exposure to dietary fibre separately by source, or colon cancers separately by subsite. These potential confounders have been known for at least 20 years. However, the disease is normally considered by epidemiologists as a single entity. More recently, it has become clear that colon cancer can arise via various histological pathways, and by various genetic pathways. There is no reason at all for assuming that risk factors for these possible pathways are the same. There is a need, therefore, for a more detailed approach to the study of diet and colon cancer, with fibre source and cancer subsite, genetic pathway and histological pathway taken into account.