Over the last three decades the condition of the teeth of children has improved tremendously. This has generally been attributed to the increased use of fluoride toothpaste. During this period the total amount of sugars disappearing into the population per capita has hardly changed. This suggests that the relationship between diet and caries has to be reassessed, which provokes different opinions among dental experts. Some suggest a maximum threshold level for the daily amount of sugars to prevent caries. Others propose that in general the amount of sugars eaten is not an important determinant of caries experience. The scientific evidence for the various opinions on the role of diet in caries development will be discussed. It is concluded that the role of diet is not so much related to the diet itself but to the individual behaviour of people. Where oral hygiene and fluoride supplementation are adequate, the diet has become a lesser factor in caries prevention. However, those diets may cause caries when there is too little fluoride. It is a mistake to classify a diet as cariogenic it may be potentially cariogenic. When one wants to decrease this potency, one should modify those factors that are actually controlling it, which is, in most cases, the topical presence of fluoride and not the composition of the diet. A model is proposed to guide caries prevention. The (insufficient) use of fluoride, (insufficient) oral hygiene and (insufficient) clearance by saliva form a window of risk. The total burden of cariogenic food that can be seen through the window constitutes the actual caries risk.