The ‘metabolic syndrome’ develops when an individual makes inappropriate responses to their environment, predisposing to obesity, dyslipideamia, Type 2 diabetes and vascular endothelial dysfunction. Whilst there is debate about the definition of the syndrome and its usefulness as a descriptor, it is associated with increased risk of coronary heart disease. Much research has focused on potential genetic determinants of the phenotypic components of the syndrome on one hand and lifestyle and nutritional factors on the other. There is however a third component of how individual risk is created, namely the interaction between genotype and environment during development. Such developmental origins of health and disease (DOHaD) accounts for a substantial fraction of risk. This is part of a broader picture in mammals, in which developmental plasticity sets many phenotypic traits during prenatal and infant life based on cues about the environment, transduced by the mother and transmitted to her offspring across the placenta or in her milk. Developmental plasticity is also utilised to produce polyphenisms in many other species too. In humans, mismatch between predicted and eventual environment can arise through unbalanced maternal diet or disease and migration or socio-economic development (1). This pathway is involved in the dramatic increases in metabolic syndrome between generations in both developing and developed societies. The effects are also exacerbated by demographic changes in reproductive behaviour, such as the tendency for women to have children at the extremes of their reproductive age and for more primiparous pregnancies. A second set of developmental pathways also exists, by which fetal or infant overnutrition affects development, to become manifest later as metabolic syndrome in the offspring. Such overnutrition may originate as maternal diabetes, maternal obesity or infant overfeeding (2). Thus the risk which starts with mismatch can be perpetuated into new cycles of risk in successive generations. Low socioeconomic status and educational attainment underpin many aspects of these cycles. The processes underlying these developmental effects involve various components of non-genomic inheritance. Particular interest concerns epigenetic processes, involving DNA methylation, histone structure and small non-coding RNAs (3). Other processes relate to parental physiology, for example maternal adaptations to pregnancy, or behaviour such as suckling and grooming of offspring (4, 5). In addition recent animal data reveal that the changes induced by dietary change or endocrine challenges in pregnancy can be passed to the grand-offspring (F2) without additional challenge in the F1 generation, and can affect a range of cardiovascular, endocrine and metabolic functions (e.g. 6). We have now shown that these effects may be due in part to epigenetic changes (7). Current emphasis on the metabolic syndrome is focused on screening and interventions in adults. We believe that understanding the epigenetic and other processes which operate early in life to determine risk holds greater hope for future detection of those individuals and population groups most susceptible and for design of appropriate interventions.