The search for the causes of CHD has been guided by a `destructive' model, which proposes that influences acting in adult life, such as smoking, obesity or high saturated fat intakes, lead to an acceleration of age-related destructive processes, including a rise in blood pressure and the formation of atheroma. One explanation for the failure of the model to account for, or indeed to prevent rising epidemics of CHD, is that individuals are heterogeneous in their responses to such influences. This heterogeneity in response is linked to different paths of early growth. The recent discovery that individuals who develop CHD grew differently from other individuals during fetal life and in childhood has led to a new `developmental' model for the disease. Reduced fetal growth followed by poor growth in infancy leads to an increased risk of development of CHD, and its associated conditions, stroke, hypertension and impaired glucose tolerance. These effects are compounded by accelerated weight gain, which may represent 'compensatory growth' in childhood.