Epidemiological studies have linked low birth weight with an increased risk of type 2 diabetes in later life. This finding has been observed in many populations worldwide, in many different ethnic and socio-economic groups. These studies led to the proposal of the 'thrifty phenotype hypothesis' that suggests that the foetal environment plays a major role in mediating this relationship. Here we review the human studies and those in animal models which support the 'thrifty phenotype hypothesis'. Molecular pathways underlying the mechanisms by which a suboptimal foetal environment leads to increased risk of type 2 diabetes are discussed, along with future directions outlining how these pathways and programming events can be further dissected to discover plausible intervention strategies to reduce type 2 diabetes.