Whilst previously type 2 diabetes occurred in older adults, its incidence, together with obesity, has increased rapidly in children. An improved understanding of this disease pathway from a developmental view point is critical. It is likely that subtle changes in dietary patterns over an extended period of time contribute to diabetes, although this type of rationale is largely ignored in animal studies aimed at determining the mechanisms involved. Small-animal studies in which large, and often extreme, changes in the diet are imposed at different stages of the life cycle can have substantial effects on fat mass and/or pancreatic functions. These responses are not representative of the much more gradual changes seen in the human population. An increasing number of studies indicate that it is growth rate per se, rather than the type of dietary intervention that determines pancreatic function during development. Epigenetic mechanisms that regulate insulin secretion by the pancreas can be re-set by more extreme changes in dietary supply in early life. The extent to which these changes may contribute to more subtle modulations in glucose homeostasis that can accompany excess fat growth in childhood remains to be established. For human subjects there is much less information as to whether specific dietary components determine disease onset. Indeed, it is highly likely that genotype has a major influence, although recent data relating early diet to physical activity and the FTO gene indicate the difficulty of establishing the relative contribution of diet and changes in body mass to diabetes.